In November 2008, Parkinson’s disease patients worldwide learned that the results of the Phase 2 clinical trial of CERE-120, Ceregene Inc.’s gene therapy approach to deliver the trophic factor neurturin directly to the brains of Parkinson’s patients, failed to demonstrate greater benefit than a placebo. The double-blind, controlled, Phase 2 trial of CERE-120 in 58 people with Parkinson’s was partially supported by The Michael J. Fox Foundation, as was the Phase 1 trial in 12 patients.

The trials tested CERE-120, a novel gene therapy product that used a viral vector to deliver neurturin, a potent nervous system trophic factor, to the brain. Neurturin is in the same family as GDNF and has shown potential in preclinical studies to slow or stop Parkinson’s disease progression.

The Foundation talked to Raymond T. Bartus, PhD, executive vice president and chief scientific officer, Ceregene, Inc., and Todd Sherer, PhD, MJFF’s vice president of research programs, about the implications for Parkinson’s patients and research directions, and for PD therapeutics development overall.

MJFF: What were the characteristics of patients enrolled in the CERE-120 study?

Raymond Bartus: These were all Parkinson’s disease patients at a so-called advanced stage. They had been diagnosed with Parkinson’s for at least five years. They had been on stable medications for some period of time, and those medications continued to do them a lot of good but were far from adequate. The patients all were experiencing what neurologists call motor fluctuations, with significant “off” periods — times during the day when, in spite of being on optimal medication, the patients were no longer getting relief from their symptoms. And despite optimal medication, many were suffering from dose-related dyskinesias. In a sense they were the exact same patients who are good candidates for deep brain stimulation (DBS).

MJFF: According to the press release issued by Ceregene about the Phase 2 results, “the trial did not demonstrate an appreciable difference between patients treated with CERE-120 versus those in the control group.” Does that mean the treatment was ineffective?

RB: Here is what we mean: When we look at the data very carefully, it’s hard to see any clear evidence that CERE-120 did the things we hoped it would do. Now, there’s no question that many patients in the CERE-120 group (those who received treatment, not a placebo, or so-called sham surgery) improved substantially from their baseline, as did many in the Phase 1 study. The problem is that an equal number of patients in the sham-treated group improved just as much. Fully 70 percent of the patients in both groups showed a five-point or greater improvement in their UPDRS motor off scores. If you are aware of the JAMA study that came out recently comparing DBS to best medical treatment, that paper defined a clinically meaningful change as five points in baseline. In our study, 70 percent of our patients showed that degree of change, while 30 percent of the patients showed less or no change and only a handful of subjects got worse. So, for both groups in our trial, 70 percent of the patients saw that same five-point, or better, change. This really makes it impossible to know if CERE-120 did any good at all and difficult to interpret the results, except to conclude we see no difference in change from baseline between CERE-120 and sham-treated patients.

While there were some signs that CERE-120 may have been doing something beyond what occurred in the sham surgery group, these effects were small and infrequent. More importantly, we missed the primary endpoint, which is the gold standard for a controlled study. For controlled studies such as ours, you must define, in advance, what the major, singular endpoint will be to determine efficacy, and then also include a number of other secondary endpoints. The primary endpoint in our study — UPDRS motor off scores — simply failed to show any difference. The two groups were equivalent.

MJFF: It is frustrating to hear that 70 percent of trial participants showed improvement, yet the study is considered a failure. How is that possible?

RB: First and foremost, all of us at Ceregene — and I know all of you at The Michael J. Fox Foundation — sympathize with patients’ disappointment and frustration. We, too, were hoping for a very different outcome.

To answer the practical question: Frankly, it’s unclear from the data whether CERE-120 is doing anything at all. While a profound placebo response like this can sometimes mask a response to the treatment, if there had been a tremendously strong CERE-120 response, we likely could have seen it above and beyond the placebo response. But it would take almost a miracle response to overcome that degree of placebo response. What we’re left uncertain about is whether CERE-120 produced even a modest clinical response that may have been obscured by the profound placebo response.

I’m not suggesting that the placebo response was the primary reason this trial failed. But it was certainly a complicating factor in determining whether there was anything going on that we could build on. In any case, however, the results of this trial must be considered disappointing, for the trial failed to show a difference between CERE-120 and sham-treated patients.

One of the things Ceregene hopes to do — and I’m hoping we can work with MJFF as well as movement disorder specialists on this — is to help try and find ways to mitigate this placebo response. It doesn’t occur in all Parkinson’s trials or even in all surgical interventions. It did not occur in Amgen’s halted 2004 GDNF trial. It did not occur that profoundly in many of the fetal tissue transplant trials. But, according to a published abstract, it did occur in the Diacrin study and apparently also recently occurred in the Spheramine trial (though this is based on hearsay information, since these data have yet to be released).

MJFF: What steps can be taken by scientists to better understand the placebo effect and, hopefully, move toward neutralizing its impact on future trials?

RB: This is important. The placebo effect is more than the “sugar pill phenomenon” that many of us think of. It’s not a simple matter of telling a patient they’re getting a miracle drug and then, in anticipating getting better, they get better.

If you look at our data, the placebo response did not occur equally in all sites. Some sites had very profound placebo responses; others showed much less. Somehow, the investigators and their staffs are playing a role in this. Those who are familiar with placebo responses understand that it’s not merely a response of the patient, but also includes expectations and possible biases of the investigators, the study coordinators (who routinely see the patients) and the raters who assess the patient’s motor performance.

We had patients who, according to their neurological assessments by the UPDRS motor off — again, our primary endpoint for trial success or failure — were doing quite well following treatment. They were blinded to treatment, of course, but we could see from the blinded UPDRS motor scores that they were performing quite well. But they didn’t feel that they were doing well.

In fact, two patients opted out of our trial to undergo DBS. They did this even though they knew that if they had gotten sham, they weren’t going to be able to get our treatment later; and if they had gotten CERE-120, they could be exposing themselves to unknown risk, because nobody had ever studied the interaction between DBS and CERE-120. In spite of the encouragement of the neurologists to stay in our trial, they opted out and got DBS. When we broke the blind, we found out that in fact both of them were on sham. They knew they were not getting better, despite scoring as if they were, during their clinical visits, on certain of the assessment scales (such as UPDRS motor off). This helps illustrate that the placebo response is not merely something that occurs in the minds of the patients, but also in the eyes of the investigators and their staff.

Todd Sherer: Additionally, the placebo response does not go across all of the outcome measures. It’s not equivalent. It’s seen in some outcome measures but not in others. That’s something that could use more understanding.

RB: You are correct, Todd, so this is not simply a matter that some patients are told that they are getting a miracle drug and therefore improve. It’s far more complicated than that. Some do indeed improve; some swear by the fact they’ve improved. Some have called MJFF and others have written about it on the Internet. Just as with the Amgen GDNF trial, emotions are very tied up in all of this; and in the GDNF trial, some who swore they improved in that trial were getting merely saline infusions. Still others are seen to improve by the raters or investigators on some measurements, but do not seem to improve on other closely related measurements or tasks.

So you can see how complicated this issue is. When you talk about a placebo response, you are actually talking about a placebo phenomenon involving many different facets and measurements that don’t necessarily even correlate with each other.

MJFF: This points to a failure in the means of measuring progress in clinical trials.

RB: That is an excellent point and one that, in my opinion, we need to get key opinion leaders and other professionals interested in treating movement disorders to buy into. There are those in our field who feel that the placebo phenomenon is not an issue; that if you have a good treatment, you can overcome it. Thus, there is hardly universal acceptance of the placebo phenomenon as a problem, even by those who are very experienced in developing drugs. I believe it is an issue that must be addressed, because developing novel, innovative therapies is difficult enough without complicating the challenge with the ‘noise’ introduced by a profound and complicated placebo phenomenon.

This problem is exacerbated, too, by the lack of a biomarker. The markers and measures we do have in PD trials are highly subjective. This is a problem in other disorders as well, such as depression. But there is a third complicating element in Parkinson’s trials: The dopaminergic aspects of movement disorders that encompass PD, themselves, play very strongly into the placebo phenomenon, because dopamine is involved with expectations and the reward system. Deficiencies in this system, compensatory changes, and pharmacological manipulation of the dopamine system may make people with PD more prone to show placebo response.

MJFF: I want to read you something that a CERE-120 trial participant wrote to our Foundation:

“…Not all the patients responded the same to the therapy. Some did quite well, like me, and some did not do as well. It is time for the pharmaceutical companies to begin looking at patients individually and attempting to determine why some responded better than others and use the resulting information to create future trials and therapies.”

Many patients share this concern. What are the next steps for the analysis of data from the trial? Will it be possible to learn why the treatment may have been more effective for some patients than others? What’s the timeframe for this analysis?

RB: We are doing some of those analyses. But the question is difficult to answer because it unfortunately starts from a scientifically invalid position. This patient is saying that some participants in the study responded to CERE-120 far better than others did. But because there was such a profound placebo response, we cannot draw that conclusion based on scientific merit. While the question takes into account that the patients who were on CERE-120 improved over their initial baseline scores following treatment, it disregards that an equal number of patients on sham responded equally well (though they received no real treatment).

Nonetheless, we did look at patients who improved from baseline in the treatment group and in the sham group. There aren’t any major conclusions emerging from that analysis. Remember, the numbers here are not that large. We have 20 sham patients and 38 CERE-120 patients.

Now, imagine a situation where we had a very weak placebo response, and we were describing the same thing — the trial didn’t work, we failed to see an appreciable difference between the CERE-120 group and the sham group — but we had a handful of patients responding really well. Under those circumstances, this would be an excellent question. You could ask why these few patients responded when most did not, and try to figure it out. That would be a wonderful scenario that you could build on going forward with further development.

But what we actually have is an equivalency in response between the two groups, with 70 percent of participants giving us a clinically meaningful response, and literally half of each group giving us a robust clinical response. This makes it difficult to parse why some may have been responding to CERE-120 better than others, because it’s obscured by the fact that so many in the placebo group responded extremely well.

On a scientific basis, we cannot conclude that CERE-120 did anything of any clinical meaning.

MJFF: Another frustrating aspect of this story is the major difference between the CERE-120 results at Phase 1 and Phase 2. Phase 1 seemed to show such exciting, robust and promising results; and then we get this extremely disappointing outcome at Phase 2. How did the Phase 2 trial compare with the Phase 1 trial that seemed to show efficacy?

RB: Certainly the Phase 1 data was very exciting and promising, as you describe it; the Phase 2 data was very disappointing but mainly because we did not see a difference between sham and CERE-120. The Phase 1 data showed about a 36-percent improvement from baseline, but there was no sham control group to compare to. While the Phase 2 showed about a 17-percent overall improvement, which was not as great as the Phase 1 trial, if the sham control group had not shown a similar improvement, the trial would not have been as nearly as disappointing and we would be having an entirely different type of conversation.

There are several likely reasons why the Phase 2 didn’t show the same magnitude of improvement as the Phase 1. One important difference is that in the open-label Phase 1 trial, 100 percent of the patients knew they were getting treated, and the investigators knew 100 percent of the patients were being treated. So anything that’s going to influence the scores, anything that’s not related to the treatment itself — that is, the so-called placebo response — will be magnified to its full extent. In a Phase 2 design, you bring in doubt, because one-third of the patients are actually receiving sham, and no one knows which ones they are. This tends to mitigate the variables influencing that placebo response on both sides — the patient and the investigator side. Therefore, it’s not surprising that you lose some of that placebo response, and that is quite common in clinical trials.

Importantly, there were many subjects in the Phase 1 trial as well as both sham and CERE-120 patients in the Phase 2 trial who showed remarkable improvement from their initial baseline score; however, the main difference between the trials was that the proportion of these patients was much larger in the Phase 1 trial than for either of the groups in the Phase 2 trial.

MJFF: How was it possible to analyze the data and come to conclusions so quickly when patients were still entering the Phase 2 trial in October?

RB: If you look at our record of conducting this program, we dosed 12 patients in Phase 1 in a matter of eight months. By any standard, this is extremely efficient execution for a ‘first in human’ trial, but is particularly noteworthy for a surgical-based gene therapy trial. We then were able to design the Phase 2 trial and gain all the approvals to launch it in less than a year from completing the Phase 1 trial. The Phase 2 trial involved testing 58 patients in nine different centers around the country and we completed enrollment in 10 and a half months, again showing very efficient execution.

To complete the data analysis of the Phase 2 in a similarly efficient fashion, we began cleaning up the blinded database long before we completed testing subjects. That allowed us to analyze the data very quickly after the last subject was seen for the last visit. That was deliberate, because we wanted to continue to execute this trial as efficiently as possible.

But at the heart of it, I think there are two reasons the data was announced as quickly as it was. First, we may have worked faster than others would have, because we feel emotionally involved and passionate about this study. Second, we felt a need to tell the world what the outcome was. We felt a commitment to share what we were doing because we knew that anticipation in the patient community was very high. Rather than sitting on the negative results for weeks or months to try to figure out how to position it or whatnot, we elected to tell the world in a matter of days. That’s been our philosophy all along. We published all the relevant papers; we wanted to share what we were learning, and we’ll publish the results of this trial, as well. I hope that other people concerned about Parkinson’s disease will appreciate and respect our ability to execute efficiently and quickly, and our willingness to share information responsibly and rapidly.

MJFF: Was dosage a factor? Is it possible that a higher dose of neurturin would have yielded different results?

RB: This is a question where different people may disagree but let me give you my opinion, based on the scientific evidence we have collected and evaluated.

Sadly, two of our study participants died from non-CERE-120-related causes during the course of this trial. This, however, gave us the opportunity to very carefully analyze these patients’ response to CERE-120 through autopsy.

We did see clear evidence that neurturin is being expressed by CERE-120 in the targeted area — that is the putamen, the terminal fields for nigrostriatal dopamine neurons. Might a higher dose have led to greater expression and potential for more pronounced effect? Absolutely. We gave the highest dose we felt we could give safely when we designed this program five years ago; we’ve learned a lot since that time, and we have a lot more safety data, both in animals and in humans than we did at that time. So there is no question that in another trial we can administer a higher dose.

But my own strong belief, and one now shared by many key leaders in the field whom we have talked to, is that increasing dose alone won’t make enough difference. Based on extensive testing in animals, we know that the dose we gave is adequate to do certain things, but it’s not doing those things in the Parkinson’s brain.

This relates to what you expect CERE-120 to do when you put it in this terminal field (putamen) that the dopamine nigral neurons project to.

When you put the CERE-120 in the terminal field of animals, the protein gets transported back to the cell body through the axons that originate in the nigra cells and project to the putamen — thus, the protein is transported from the putamen back to the nigra. This is important, because in order for neurturin to repair these neurons, make them function better and protect them from further damage, it has to get to the cell body of the dying or damaged neurons. Indeed, in all the pre-clinical studies, we not only had evidence of efficacy, but we saw evidence of neurturin back in those cell bodies. However, in the two Parkinson’s brains, despite clear evidence of neuturin in the putamen (where we administered CERE-120), we saw no evidence of neurturin in the cell bodies (where it required transportation via the axons of those cells).

Based on all the pre-clinical and autopsy data we have accumulated in this program, the problem is clear. Delivering CERE-120 to the terminal field (putamen), only, is sufficient in pre-clinical models, but not in the brains of people with Parkinson’s. This suggests a serious problem with axonal transporter mechanisms in Parkinson’s disease. This is a concept that’s growing in popularity and understanding. It is much more clearly established in other diseases involving neurodegeneration, such as ALS and, to some extent, Alzheimer’s disease, but it is gaining acceptance in the movement disorders and Parkinson’s fields. The concept is that, long before the cells die, there’s a loss of processes involving the axonal terminals; and that even before the loss of axonal terminals, there is a deficiency in transport ability within those terminals. While it is our best hypothesis that transport mechanisms in Parkinson’s disease are severely impaired, it is a fact that in the two patients we examined, we saw no evidence for neuturin in the nigra (where we needed it to get), despite adequate levels expressed in the terminal field (where we delivered CERE-120).

Additional support for this interpretation comes from further analysis of the human tissue and its comparison to our pre-clinical experiments. As I stated earlier, only when neurturin gets to the cell body can you expect it to begin repairing the neuron and making dopamine synthesis and other important activities more effective. One way you measure that is by looking at tyrosine hydroxylase (TH), the major synthesizing enzyme for dopamine. In all the pre-clinical studies, when you get CERE-120 in the terminal field, you not only see neurturin expressed in the terminal field and in the nigral cell body, but you also see enhanced TH in both the terminal field and in the cell body. That shows us that when you put neurturin in the terminal field and it gets carried back to the cell body, the neurons become more healthy and functional, including as a better dopamine machine. We saw no evidence for any of that in the Parkinson’s brain. Despite a good neurturin signal in the terminal field, we saw none of the secondary consequences of the neurturin expression that are required for improved function and neuronal repair that were consistently seen in the animal studies. So my point is, simply adding more protein to the terminal field likely won’t do anything because that will not ensure it also gets back to the nigral cell bodies, as is required for efficacy.

MJFF: It’s not getting from Point A to Point B. If the car is broken down, putting more gas in the tank is not going to help.

RB: Exactly. Now, we do have a proposed solution to this, and that is to also target the nigra. That is, in addition to targeting the putamen, we hope to revise our dosing strategy to target the nigral cell bodies as well.

We’ve been all over this idea for the past several weeks. We’ve talked with neurologists, neurosurgeons, psychiatrists, neuroscientists and other key opinion leaders about targeting the nigra. It turns out that a lot of neurosurgeons target the nigra indirectly during deep brain stimulation (DBS). This is because the subthalamic nucleus is stimulated with DBS, and is located very close to the nigra. The surgeons actually go through the subthalamic nucleus to the nigra to confirm where the needle is, and then pull back to the subthalamic nucleus again. So they’re familiar with locating this precise area of the brain.

The question is what kinds of doubts might exist to doing it. In terms of side effects from CERE-120, we’ve talked to a number of opinion leaders and we suspect that this can be done safely and effectively.

This change in dosing strategy would require us to go back to discuss with the FDA, which of course is part of our plan.

Nonetheless, even if we were to go forward in targeting the nigra, it’s important to note that the placebo phenomenon would still be a factor in that study. It is possible that we could do a second trial and still not get a result distinguishable from placebo effect.

MJFF: If you launch a second CERE-120 trial, will you be able to re-enroll the trial participants who received sham treatment in the first study?

RB: I actually raised that issue and discussed it with my colleagues. Wouldn’t that be a great thing to do? You have these patients who were courageous enough to volunteer for this study, but they got sham treatment, so wouldn’t they therefore be great candidates? From a humanistic standpoint it would be a great thing to do. But from a scientific standpoint, you quickly realize that you would really compromise the program by going in that direction.

First of all, the patients have now progressed another year or two from where they started with our trial. There is a very general consensus, based on some data and a lot of intuition, that the more the disease progresses, the less benefit one might expect from CERE-120. Thus, using these patients for an important ‘go/no-go’ decision could represent a serious mistake if the dosing change indeed worked but we did not see a strong effect because of their disease progression.

Secondly, 70 percent of those patients have already showed a robust response — that is, a placebo response. If they already have an elevated baseline, we could be shooting ourselves in the foot trying to overcome that. You see how complicated it gets.

MJFF: What about the patients in the group that received CERE-120? Could you re-enroll them?

RB: Unfortunately not. From a scientific validity perspective, the findings would be compromised by working with patients who had already received an earlier form of the treatment. In order to be sure that any results you see are coming from the treatment you are currently testing, you would have to work with patients who had never received any form of CERE-120.

Moreover, there are significant theoretical safety issues with dosing the same patient a second time with the AAV viral vector we use to deliver the gene for neurturin. We have not yet performed all the necessary animal studies to demonstrate the safety of multiple AAV injections (our program currently assumes that one injection is sufficient for the lifetime of the patient). It would take a couple of years to complete those studies before we could even begin to think about dosing the same patients with CERE-120 a second time.

MJFF: Do you think it would be a challenge to find people to enroll in a second trial?

RB: We’ve talked to a number of opinion leaders about our concept to add the nigra as a target. In so doing, we included a lot of the principal investigators and surgeons who participated in our Phase 2 trial. Most, if not all, of them were very enthusiastic about it and do not think they would have trouble recruiting patients into the trial. Of course, it remains to be seen whether that’s the case, but that’s the feeling right now.

Most of the investigators, frankly, like us and MJFF and the patient community, were worried that the negative trial outcome would just stop everything. I must say, we’ve received a great deal of encouragement from these investigators, saying the program still looks scientifically sound, it looks like one of the best things going on, I hope you guys don’t give up.

MJFF: For the trial participants who received CERE-120, what comes next? CERE-120 will continue to express neurturin in their brains for the rest of their lives.

RB: You are correct: Everyone in the CERE-120 group will have the treatment present in their brains for the rest of their lives. That is a major difference from the situation that existed with Amgen and GDNF and the pumps and the hardware. The sites we recruited to enroll these patients will try to continue following up with them routinely so that we can be certain they are doing well, and so that any long-term side effects that may emerge are documented and treated.

Todd Sherer: This may be something our Foundation takes into consideration as an aspect of the possible value of keeping the study going.

MJFF: Given the differences between the pre-clinical data and the results of the human trials, do you think there will be implications for how pre-clinical trophic factor research is conducted going forward?

RB: There could be. Let’s face it: There’s no animal model that recapitulates any human disease completely, and it is a significant problem in Parkinson’s disease research. We’ve known that, with a trophic factor approach — where you’re not merely trying to modulate activity in the terminal area per se, but are trying to restore the vitality and function of the entire cell, which means affecting things at the cell body — you have to make sure that the trophic factor gets to the cell body. Our data in the two autopsy cases shows us that, contrary to what happens in the animal models, CERE-120 delivered to the putamen does not result in neurturin expression in the nigra. Thus, they offer further support for the hypothesis that one of the problems in PD, which leads to dopaminergic motor symptoms, is a loss in synaptic function includes axonal transport and this specific deficiency has thus not been the focus of animal models for PD.

In terms of basic research, if people could understand the mechanism that details the loss of synaptic transport — confirm its prevalence, or what’s occurring, and try to understand it better — then that might be something you could build on. You could possibly try to have drugs that aim to enhance that function. A second thing would be to try to develop better models that mimic longer-term degeneration of the axons. MPTP is very quick. 6-Hydroxy dopamine is very quick. And it goes so quickly that you can’t study a snapshot of the degeneration; nobody has modeled the slow degeneration aspect of the disease. And to the extent that this is an important pathogenic event, better models that capture that would lead to improvements in pharmaceutical testing.

From our standpoint, I think we can get around the lack of animal models by simply targeting the nigra. But certainly, if we had known more about this issue prior to launching our clinical program (for example, if this element of the Parkinson’s field had advanced 10 years earlier), we wouldn’t be in this position, because we would have recognized a need to target the nigra right from the start.

MJFF: Do your results change the way researchers think about neurturin, or trophic factors overall, as a potential treatment for Parkinson’s disease? Will trophic factor research continue?

RB: When you have something as disappointing as this is, it leads to a lot of different questions. We’ve done soul-searching; we’ve tried to be very objective, but who knows? Many of us at Ceregene have put a lot into this program; we’d like to see it go forward. And while we’ve tried to do reality checks, and we talk to a lot of outside people so that we don’t engage in a bunker mentality in our own thinking, and we have obtained widespread encouragement for continuing our effort, even those people may have positive biases toward this approach generally, and the CERE-120 program specifically.

I believe that what this trial demonstrated is greater difficulty in using the trophic factors than we imagined, and the need for more extensive targeting, as I’ve described. At the same time, I admit that other people may look at it other ways. Some may say: ‘Wow. This is one more failed trial with trophic factors, and like all the other failures this was a delivery issue, and this is just one too many. Nobody’s ever going to solve this thing.’ And while it is possible they may be right. I don’t think they are. I believe that if you target the nigra, along with the putamen, it may provide the same trophic response observed for decades in pre-clinical studies. This conclusion is consistent with the data we now have; we’ve always recognized that you must have the trophic factor protein in the cell body, but did not recognize that placing it in the terminal field would not be sufficient to achieve that end in Parkinson’s patients (as occurs in the pre-clinical models).

And it might be that by the time you reach an advanced stage of Parkinson’s disease, the status of those neurons, and the fibers that extend to the putamen, may be so degenerated that it’s impossible for any nerve growth factor to restore them. That’s the $64-million risk that exists going forward. And that’s just a logical possibility. It also argues for the possibility that the biological phenomenon that you see with trophic factors in animals doesn’t transfer to neurodegenerative diseases in humans. Because there’s not a single animal that has all the constellation of issues and problems that encompass what we call Parkinson’s disease. And therefore the totally different issues that exist in what we call Parkinson’s disease may be too much to overcome with any one single approach. These are all very plausible positions that people could infer or conclude from the results of this trial.

The encouraging scenario, though, is the one I’ve already described to you. And it’s the one that I embrace. I continue to look to the future with a great deal of hope.

# # #