Reprinting an email from Greg Wasson with his permission regarding another treatment cancellation.

Has it occurred to anyone that there may be an important pattern here? In the GDNF trials, in the trials recently discontinued with Spheramine, in fact in the forty years since the development of the gold standard of carbidopa/levodopa, there have been a remarkable number of trials that appeared to have dramatic results on patients in early clinical stages but which failed to meet statistical endpoints in placebo-controlled double-blind phase 2 trials.

Could anyone who had seen Peggy Willocks in 2000 before her surgery doubt in any way that her surgery had a dramatic and lasting positive effect on her ability to function. Does anyone doubt that the film some of the GDNF clinical trial participants before and after receiving that biologic showed clearly that some of them responded in a way which could not be attributable to a placebo response?

Could it possibly be that some patients are legitimately responding in a
dramatic fashion to these experimental therapies, while others are not, regardless of the presence of placebo controls? Is it possible that the reason for these apparently topsy-turvy, now you see it now you don't results are themselves the result of clinical trials which are fundamentally flawed in their design; that is, if "idiopathic Parkinson's disease" is in fact a bundle of related diseases as has been increasingly discussed, and may well be a disease, or diseases, which affects many more parts of the brain than had traditionally been assumed, it may well be that it's not the data that is wrong in recent clinical trials - it is the square peg of data which the scientists are trying to fit the into what they refuse to recognize as the round hole of current clinical trial design.

A child can see that the symptoms from patient to patient very so dramatically that one would intuitively assume that we are looking at different diseases. How do we make the leap to the assumption that despite the evidence of our eyes, we all have (with some exceptions) "idiopathic Parkinson's disease"?

It is rather ironic that this so-called "designer disease" can only be diagnosed through clinical observation. Clinical observation alone would lead the average person to assume that more than one disease, or form of the disease, is present in the population. But since the discovery that carbidopa/levodopa dramatically ameliorates the symptoms of PD in almost all patients regardless of their presenting symptoms, and because we can detect a dramatic loss of dopaminergic cells in "PD" patients through brain scans and at autopsy, we leap to the conclusion that we all suffer from the same exact malady.

It seems to me that the existence of several variants on what we call one disease would explain the varying results or effects on patients with regard to new experimental therapies. We know so little about the disease in the larger sense that we even call it "idiopathic," meaning of unknown origin. Remember, only a few short years ago, even a movement disorder specialist would have told you that Parkinson's, in the often repeated comparison, was the disease that left your mind clear while trapping you in a body that was eventually completely immobilized, while Alzheimer's was the disease that destroyed your mind while leaving your body intact. I don't think that any movement disorder specialist worth his or her credentials would make that same statement todayabout Parkinson's disease. We know that it affects to some degree in almost all patients the frontal cortex. And it appears to affect other areas of the brain
as well. Could it be that we have been so mesmerized for the last 200 years by the massive and dramatically observable destruction that Parkinson's does to the portion of the brain devoted to movement that we failed to recognize what in hindsight should have been obvious secondary symptoms of the disease or set of related diseases?

We all know that a new paradigm of Parkinson's disease is emerging that in some ways is consistent with the view of Parkinson's that I have outlined above. What we will finally learn about Parkinson's and its causes, and the dramatic variance of symptom constellations among individual patients, remains to be seen. But I think that it's time to take a good hard look at clinical trials and their results in light of this emerging new view of what we had assumed was a single disease. If Parkinson's is indeed a cluster of related diseases it would make sense that some therapeutic applications would work among all or nearly all of the patient population, such as in most, but not all cases, carbidopa/levodopa. It would also make sense that certain dopamine agonists would work among a large portion of the population, but leave a significant minority of patients unaffected or with a negative response to the therapy.

Perhaps most importantly, it would explain why for year after year, decade after decade, we have been getting to large phase 2 trials and winding up with statistical endpoints that may in fact be unachievable for most therapeutic applications because only certain patients with certain variants of the disease are going to respond. It may be that bad trial management in some cases (disconnected catheters and poor infusion techniques in the GDNF trials) and placebo effect itself are "noise" that interferes with a proper interpretation of what these drugs are doing to patients. A proper interpretation of this pattern of clinical trial failures may mean having to conclude that the trial design itself is based on the false premise that we all have the same variant of the same disease.

As Andy Grove has posited, why aren't we learning from our failures? Why do we keep coming at the same problem from the same position with the same set of assumptions? Why don't we at least entertain the notion that the way we test new therapies is inconsistent with the disease, or diseases, that we are trying to cure. In fact, one would almost think that the refusal to budge from the current thinking about Parkinson's clinical trials reflects some of the cognitive damage that Parkinson's does in the tendency of patients to exhibit a stubborn inability to move from one approach to another. All the while, we may be throwing away the baby with the bathwater, as therapy after therapy that may in fact address major problems for some of the patient population, are deemed to be failures and thrown away. If so, it is a mistake of gigantic significance for us all.

I hope I haven't stated the obvious, or showed my ignorance of clinical trial design and biomedical science, but when I advanced this idea at a PD meeting two weeks ago, I looked across at an English neuroscientist who nodded his head "yes" the entire time I was speaking. Interestingly, however, he remained silent.

Just thinking out loud,

Greg

Here is what Andy Grove said in his speech last Spring to the Society for Neuroscience:

The Holy Grail would be a biological measurement of progression that is objective, repeatable and rock solid. We don't have that. Partly we don't have that because it's complicated. Partly we don’t have that because in my opinion we don't work hard enough in proportion to the complication of the task.

and his metaphoric reference to the “gold nugget” :

The most important characteristic of failure is that it is a golden nugget, something that one should polish, put under a magnifying glass, analyzed --because we don't know what you're going to find when you take the dirt off.

I am working on an OpEd or letter to the Editor or something – I cannot just ignore this opportunity! I lucked out on getting a letter published in the WSJ last time – this one will be more difficult.

Peg

I just finished Nick Nelson's fine book about the GDNF trials by Amgen called "Monkeys in the Middle". As a medical researcher, I can tell you that whoever designed the Amgen trials didn't know some of the basic scientific aspects of organizing and running such a study. I worry that the same may have been true for the Spheramine trial and hope that the Ceregene and Neurologix trials have better research design consultation.

For the Amgen GDNF trials it appears that the various treatment sites had a great deal of latitude in the tubing that they used to pump the GDNF into the brain and the way in which this was accomplished. Some of the pumping strategies were better than others and there were site specific differences in outcomes. The other very distressing issue is that in many of these studies the major outcome measure is the Unified Parkinson's Disease Rating Scale (UPDRS). The truth of the matter is that the UPDRS is a lousy measure with poor inter-rater reliability. It is highly subjective with poor guidelines for raters to use in measuring what they are seeing.

According to Nelson there is 25% or more variance between raters in scoring the scale. If the endpoint you are looking for to deem a treatment as effective is 30% improvement from time 1 to time 2, a finding or a lack of finding may be due totally to a lack of rater reliability from one time to the next.

I have been doing research on substance abuse and pain for about 20 years and although that area has some similar problems, they are not as bad as the ones that I have seen in the Parkinson's area.

Nelson's very strong conclusion is that for many people from some of the GDNF sites the outcomes were outstanding and irrefutable some sites did not get such great results. There also looked to be some systematic differences between sites that could account for these differences.

On a hopeful note it appears that within Amgen there came to be a growing feeling that the pump was not a good long term solution to the delivery of GDNF to the brain and that the way to go was gene therapy. All of us hope that this will be borne out when Ceregene reports their Phase II results in Q4. Just pray that they ran the study carefully and in a way that will show differences between treatment and control if they exist.

You couldn't be more right about the UPDRS, but it the only one field tested and acceptable at the present time. This instrument of measurement used in most every PD trial, BUT trials should allow for the variance you discussed.

The unreliability of the UPDRS has been going on for years (I could find documentation back in 1994).

http://www.ncbi.nlm.nih.gov/pubmed/8...gdbfrom=pubmed

And this statement was made in 2002!

Although the notion of placebo effect often implies patient-based perceptions, we found subjective changes to be infrequent in placebo-treated patients, suggesting that either: (1) the placebo effect was rater-driven; (2) the ADL questionnaire is insensitive to transient but objectively demonstrable motor changes; or (3) that the objective changes, albeit major, are within the realm of natural variation in the UPDRS motor scale from visit to visit.

http://www.ncbi.nlm.nih.gov/pubmed/1...gdbfrom=pubmed

Currently, there are some studies about the UPDRS such as:

1) a comparison of "on" as described in patient diaries and the UPDRS,

http://www.ncbi.nlm.nih.gov/pubmed/1...ubmed_RVDocSum

2) a new measurement tool is in the making - the UPSI (United Pasrkinson's Symptom Inventory) and in comparison to the UPDRS is proving valuable,

http://www.ncbi.nlm.nih.gov/pubmed/1...ubmed_RVDocSum

3) A recommendation was made by the Movement Disorder Society (MDS) in 2004:

Strengths of the UPDRS include its wide utilization, its application across the clinical spectrum of PD, its nearly comprehensive coverage of motor symptoms, and its clinimetric properties, including reliability and validity. Weaknesses include several ambiguities in the written text, inadequate instructions for raters, some metric flaws, and the absence of screening questions on several important non-motor aspects of PD. The Task Force recommends that the MDS sponsor the development of a new version of the UPDRS and encourage efforts to establish its clinimetric properties, especially addressing the need to define a Minimal Clinically Relevant Difference and a Minimal Clinically Relevant Incremental Difference, as well as testing its correlation with the current UPDRS.

http://www.ncbi.nlm.nih.gov/pubmed/1...gdbfrom=pubmed

4) The MDS went on record in 2007 with recommended changes :

This article presents the revision process, major innovations, and clinimetric testing program for the Movement Disorder Society (MDS)-sponsored revision of the Unified Parkinson's Disease Rating Scale (UPDRS), known as the MDS-UPDRS. The UPDRS is the most widely used scale for the clinical study of Parkinson's disease (PD). The MDS previously organized a critique of the UPDRS, which cited many strengths, but recommended revision of the scale to accommodate new advances and to resolve problematic areas.

http://www.ncbi.nlm.nih.gov/pubmed/1...ubmed_RVDocSum

If any of you are involved in making these changes of measurement, be familiar with this information here.

Peggy

Pegleg,

This information is enormously helpful. Our research group took a scale that had been used extensively in substance abuse research, but which was tremendously unreliable and changed it dramatically into a computer administered measure that has excellent psychometric characteristics. It is now very widely used.

Perhaps something similar could be done in this area. I need to think about it. But your links are a great start. When your measurement tool has more variability than your target outcome, how can you possibly be able to determine whether a given therapy is working?

well, as somebody I read recently said - forgive me if I'm repeating info that just appeared on this forum - I'm getting a little cyber-surf-dazed - but the standard testing for Parkinsoln's symptoms occurs in one 15-minute time period wherein a patient is evaluated for movement, affect, walking, speech, and lots of other things. Considering how much our conditions can fluctuate even depending on who is in the room, it seems that results could vary significantly...

From an old thread:

http://neurotalk.psychcentral.com/thread44248.html

This article about the failure of MitoQ and CoQQ10:

http://www.medpagetoday.com/MeetingC...eeting/tb/9205

"Richard Newman, M.D., a practicing neurologist in Rockledge, Fla., and discussant at the session, said it was too soon to give up entirely on anti-oxidants.

"Parkinson's disease isn't one disease," he said, but rather a collection of similar syndromes with different causes.

Dr. Newman recommended additional studies in certain patient subgroups -- such as those whose disease is related to exposure to the insecticide rotenone or to Agent Orange, the Vietnam War-era defoliant -- in whom oxidative damage may play an especially strong role.

But Dr. Snow said such studies were unlikely to be funded. In commercial drug development, "you usually only get one chance" to show efficacy, he said. "

**********
Another example of how choosing patients matters...

From this thread:

http://neurotalk.psychcentral.com/thread29755.html

TOPIC: Mixed lineage kinase inhibitor CEP-1347 fails to delay disability in early PD

I wrote:

Thanks to all. But were you tested for LRRK2 mutations prior to the study? Specifically the G2019S? It doesn't make a lot of sense to choose patients with different pathology and genes, especially for a targeted drug like this one.

"The G2019S substitution is situated in the mitogen-activated kinase kinase kinase domain of Lrrk2, within the hinge region of the highly conserved activation segment: Asp-Phe-Gly (DFG).[3] Based on in silico modeling using a series of homologous mixed-lineage kinases as templates, we hypothesize that the mutant form may modulate kinase activity or alter substrate specificity (Fig 2). Notably, Lrrk2 I2012T, G2019S, and I2020T substitutions, which cluster in this domain, also create potential novel targets for phosphorylation. Although the signaling or secondary messenger phosphorylation cascade has yet to be elucidated, we propose Lrrk2 kinase activity may promote -synuclein pathology either directly or indirectly via a protective response. Work in Drosophila melanogaster suggests -synuclein phosphorylation at amino acid S129 may be a critical event in neurotoxicity.[20][21] It remains to be shown whether -synuclein phosphorylation is a direct or downstream target of Lrrk2 mitogen-activated kinase kinase kinase activity; nonetheless, we postulate that Lrrk2 substrates play a crucial role in PD pathogenesis, and the amount of these substrates within the cell may have a major impact on G2019S disease parameters such as age of onset.

Interestingly, phase II/III neuroprotectant therapy for PD[22] with a mixed-lineage mitogen-activated kinase kinase kinase inhibitor (CEP-1347)[23] was initiated by the Parkinson Study Group, but recently was terminated due to ineffectiveness in the interim evaluation. In future clinical studies of mixed-lineage kinase inhibitors, it may be useful to stratify PD patients based on the presence of the G2019S carrier state. "

You think??

From:

Lrrk2 and Lewy body disease

http://www3.interscience.wiley.com/c...RETRY=1&:cool:

**********

Autopsies of people with LRKK2 mutations show very different pathologies, even within the same family. PD is NOT one disease. So how to choose a cohort that makes sense? They need to figure this out before wasting more money on failed trials. It's as if they don't want to bother doing it right. Heads in sand.

~Zucchini

Dr. Newman is my neurologist. How nice to see him quoted - he's a man who isn't afraid to learn from a patient.

paula

Wow, Paula, you're lucky to have a neuro like that. Did he ever give you his opinion about "banding"?

~Zucchini

ZF,

He only said that he had heard of it when I appeared banded from head to toe at an appointment looking pretty funny. He didn't offer an opinion. Without fail he asks about GDNF.

paula