Frederick Hawthorne and the history of boron chemistry go hand in hand. Hawthorne helped create the field in the 1950s, and has done more than anyone to show the fifth element’s utility, for everything from rocket fuel to pharmaceuticals. Along the way he has garnered nearly every accolade a scientist can, including the 2003 King Faisal Award. It is medicine — and cancer treatment specifically — that has Hawthorne embarking now on a new venture in collaboration with his employer, UCLA. Together Hawthorne and the University are creating a foundation to develop an obscure, but potentially revolutionary way to treat tumors. Known as Boron Neutron Capture Therapy (BNCT), it puts to use decades-old theory to selectively destroy cancer through small nuclear explosions at the cellular level.Science Blog spoke recently with Hawthorne about boron, and about BNCT’s potential and maddening history over the last 50 years.
What is it about boron in the medical setting that has kept you engaged for so long?
I first encountered boron chemistry in 1956, working on rocket fuels for the U.S. Army and Air Force. Certain boron compounds were candidate rocket propellants and therefore the research I was doing at the time in industry required boron chemistry. One thing led to another until eventually my academic career became primarily focused on boron clusters, and I’ve been instrumental in developing that field.
The thing I encountered medically early on was the possible role of boron chemistry in the cure of cancer. This took the form of the phenomenon known as boron neutron capture therapy, in which the boron 10 isotope, one of the two isotopes of boron, reacts with a neutron and forms lithium 7 and an alpha particle. When they’re generated, they form with the accompaniment of about 2.4 eV of kinetic energy plus a gamma photon. So in essence it’s an energetic nuclear fission process.
If this nuclear reaction can be harnessed or localized by having the boron pre-placed inside a cancer cell and that site then radiated with neutrons, eventually a fission reaction will occur generating these energetic particles which kill the cell in which the boron10 isotope resided.
The nice thing about this is that the particles produced during the fission reaction have a very short range. They only have a path of about one cell diameter before they’re extinguished and they become innocuous. So this is a very localized and a very deadly — in a good way — process. It can be, we believe, selectively harnessed by placing the boron 10 isotope in cancer cells.
There are many schemes afoot to do this, and some work better than others. What we know is that it’s possible to gain selectivity ratios of 5, 10, even as high as 20-to-1, of cancer being favored over normal tissue. In practice, a ratio of 3-to-1 would be of therapeutic interest.
That’s one medical aspect of boron that I’ve been intrigued by. Other interests involve the use of boron as a surrogate for carbon in pharmaceutical design and in the design of vehicles for drug delivery.
BNCT isn’t new. What’s the urgency now?
BNCT was first proposed by Locher back in the middle 1930s, soon after the identification and characterization of the neutron by Chadwick in 1932. Locher, a physicist at Swarthmore College, published a paper in about 1936 in which he proposed the boron neutron capture reaction as a therapeutic device for the selective destruction of cells. That was the intellectual birth of boron neutron capture therapy. The study of BNCT really got underway following World War II, when neutrons became readily available from nuclear reactors. The idea of using BNCT for cancer therapy was initiated by William Sweet at Massachusetts General Hospital around 1952.
Dr. Sweet was a neurosurgeon and the great progenitor of BNCT, which he started with a bang. He began early-on with a human study of terminal brain cancer at Brookhaven National Laboratory. The most insurmountable problem involving neurosurgery at that time — and still is, actually — was finding a cure for glioblastoma multiforme. That’s a very deadly form of brain cancer that usually kills people within twelve months of their being diagnosed.
Sweet set out to find a cure, and thought that BNCT would be a dandy tool to have available.
The clinical trial Sweet subsequently ran at Brookhaven was on terminal patients. It was unsuccessful, for three main reasons. One, the inavailability of suitable boron compounds to serve as target molecules or ions. The second problem was the lack of a clean and well-characterized neutron beam. They had a neutron beam from the Brookhaven medical reactor, but it proved to be very dirty by today?s standards. It had a lot of gamma and fast neutrons in it. The third problem was the lack of knowledge of the interaction of neutrons and high-energy fission products with living cells and tissue.
This is the area known as dosimetry. The dosimetry data they had available were minimal.
In a nutshell, patients were overexposed, compounds were non-selective and the beam was dirty, so they all died. It was not a good outcome.
That didn’t stop the idea, because about that same time there was a tremendous burst of activity in boron chemistry, which I was fortunate enough to lead at the time. This dealt with the discovery and exploitation of several large families of molecules and ions containing boron that were based on clusters, or cage molecules, in which the vertices of the cages were boron atoms. This opened up an area of inorganic chemistry which was a rival to organic chemistry. It lead to the development of a number of biologically safe agents which had very high boron contents, which is just what is needed for BNCT.
Armed with that new chemistry, the neurosurgeons came back into business in the 1960s with another Brookhaven clinical trial using these new molecules. Again, the reactor wasn’t much better. The neutron beam wasn’t terribly good, and the dosimetry problem still existed, and indeed, the chemistry problem still existed too, because a sufficient search had yet to be done in boron chemistry to find safe and selective molecules for targets. So that particular study also failed. The patients all died. Again, the only clinical trial ever considered was for brain cancer.
No one else proposed using BNCT for other cancers?
Well some of us were certainly curious about tumors that were killing our friends, neighbors and family members; people dying of common forms of cancer. It was certainly on our minds, and it was evident that the applications proposed for BNCT were quite biased towards the interests of the neurosurgeons, not to oncology as a whole. This particular attitude and course of events has persisted up to the present time.
Is it that glioblastoma is not a good candidate for BNCT or simply that it’s hogged all the study resources?
The answer is yes to both of your questions. Glioblastoma multiforme probably is not a good candidate for proof-of-principle. The problem is that with BNCT, no matter where it’s used, a large amount of boron must be delivered to the tumor. This is large compared to the quantity of drugs normally delivered to tumors, say for chemotherapy.
Glioblastoma multiforme is buried back in the head in the active brain — which one needs to function, to keep the body going. It’s also diffuse; that particular tumor sends out fibrils, or microscopic colonies of active cells which later become much bigger, and lead to the progression of the disease. So to do it right, you have to find and kill all those microcolonies.
Also, the brain itself is protected from blood by the blood brain barrier. And blood is the principal way to deliver drugs. So you inject a compound into the blood and it circulates, it reaches the brain and the compound must cross the blood brain barrier and diffuse through the normal brain to find the tumor and be selectively taken up in large quantities.
One way around this is that at the active tumor site, the blood brain barrier is often destroyed, so you have direct contact of the chemical in the blood with the tumor. But that’s only at that site. These little fibrils, these little microcolonies, are well separated and the blood brain barrier is protecting them. So this is really a problem that in my opinion makes glioblastoma multiforme tumor a poor primary target for proof-of-principle.
What happened after the second trial?
There was a retreat and people began to worry more about developing a cleaner reactor. Engineers and physicists were doing that. Chemists went back and tried to come up with more selective delivery methods. That succeeded, in part.
Dosimetry people had not had good computers. But with the advent of newer machines, they now had really excellent means of modeling medical phenomena: drug delivery, radiation delivery and the interaction of radiation with tissue.
All that was developed and yet no one really had a boron target compound. There were actually hundreds of viable candidates synthesized, each having its own potential usefulness for selective delivery, but very few of these were ever evaluated. Consequently, two drugs came out of the early clinical trial in the 1960s: BSH, which is about 50 percent boron by weight; and paradihydroxyboronophenylalanine, a boronic acid which only contains about 4 percent boron by weight. Only these two materials were approved by the FDA for human use. The newer compounds, which the chemists made and thought were promising, were never evaluated using large animals. They usually stopped at the mouse stage of evaluation.
The part of the BNCT program which was sheltered under the clinical banner was sort of huddled around the concept of running another clinical trial as soon as they could with cleaner reactors and better dosimetry — but they had to use the old compounds that came out of the 1960s and 1950s.
This happened again in the 1990s. The early 1990s saw a third Brookhaven clinical trial of glioblastoma, using the phenylalanine derivative, which is called BPA. It utilized about 50 terminal patients and lasted four or five years. The conclusion from that study is that BNCT with that compound, a very good reactor and good dosimetry had no more efficacy than conventional photon radiation for the same tumor. The difference was the quality of life. BNCT is a one-shot treatment, where with photons you’ve got to continually endure more and different fractions of radiation. And your hair falls out and so on. But with BNCT, patients were happier, although they still died in the same period of time, about a year from diagnosis. A long-term case would be 18 months from diagnosis.
Immediately following the second Brookhaven study, and perhaps what kept things alive through the 1970s and 80s, was the work of a Japanese physician named Hatanaka, a neurosurgeon who came to the U.S. in the 1960s, studied under Sweet and took the compound BSH back to Japan with him.
Hatanaka began his own clinical trials using his own crude reactor and the BSH material. Dosimetry was very primitive, everything he did was necessarily not very sophisticated because of lack of resources. He eventually treated something like 150 people. He died about 10 years ago. His legacy was that he kept BNCT alive through these continuous clinical trials he was running in Japan. So on the one hand, he kept pushing BNCT in the direction of brain tumors, but on the other, if not for him, BNCT might not have made it this far. Sort of a good news, bad news situation.
A reexamination of all his data reached the same conclusion that was reached a few years later in this last Brookhaven study: BNCT for glioblastoma multiforme is no more effective than conventional photon therapy. This again points to the fact that glioblastoma multiforme is not the disease of choice to investigate using something new like BNCT.
You’re not going to win. Again, no one did anything with other tumors, except melanoma. Melanoma trials were run by Mishima in Japan, beginning in the late 1980s, using the BPA molecule as a target. The difficulty with melanoma is that if you can see it on your skin, you can certainly remove it surgically. You don’t need to burn it off with neutrons and boron. The sneaky bad part of melanoma is that metastasis is so bad — it goes everywhere and it is very aggressive. You can’t really cure a person unless you can get the metastases if they exist. If they don’t exist, it means it’s a young tumor, and surgical methods are effective for that.
With so many failures along the way, what is it about BNCT that makes you optimistic that it can be effective?
I think you want to look at diseases that are very plentiful and more meaningful in terms of populations. Lung cancer, for example. We have designed a protocol for lung cancer therapy based on the use of global boron compounds and fast neutrons. This is a different set of reactants than conventional BNCT, but it’s the same principle. That looks pretty good and it is something that should be explored. Cancer of the prostate and breast cancer are also candidates which with the right compounds might be eradicated using BNCT.
Is it because of promising results in animal studies that you believe in this?
That’s kept some of the target compounds and interest alive. But even that’s limited. Large animal studies, which you need as a prerequisite to human studies, have been minimal.
Where do you go from here?
What I’d like to do is go back and find a means to revive interest in the lung cancer study with fast neutrons, the work that was all set up and almost ready to go. The human biodistribution work had been done, the toxicity work had been done, everything was ready to go. But the last necessary increment of funding to support the study was withdrawn for non-scientific reasons, and so everything that had been worked on up to that point was still good but never allowed to go to fruition with a human clinical trial.
So that would be one place to go back and begin.
Another thing to do would be to look at head and neck tumors. We have developed over the years a delivery system for boron compounds based upon liposomes. Liposomes are very tiny particles, hollow particles, which are filled with a solution of a boron compound suitable for BNCT targeting. A liposome is about 1/1000th the diameter of a cell, so it’s really quite tiny. The liposomes we have are self-selective. When injected in an animal such as a mouse or a rat with a tumor, the liposome is preferentially taken up by the cancer cells in the tumor. There we get selectivities of 3, 5 and even in some certain types of animal tumors, as high as 20.
You believe you can achieve that same selectivity in human cells?
We hope so, yes. We’ve got to run biodistribution studies in humans and find out. But we have to do more large animal studies first.
What do in vitro studies show?
In vitro studies with liposomes, that’s not going to tell you the whole story. And it seldom does with any compound. The trouble is that many agents that are used to deliver boron become attached to the external part of cells, and are physically absorbed into that cell if they are grown in a culture media. So you get a lot of boron on the surface, and if you irradiate those cells with neutrons, you’re going to wipe cells out like mad, and think, “Boy, I’ve really got a winner here.”
However, in vivo the same kind of cells and the same agent do not behave in the same fashion. Consequently, you don’t know if you have anything of value until you look at it in an animal study.
The other thing that’s important is the biodistribution of the boron compound. When injected in an animal with a tumor, there’s a competition, normally, between the tumor and the animal’s liver, kidney, and spleen. Those are the four big contenders trying to take the boron compound out of the bloodstream. With in vitro studies, you don’t see that. You need the animal studies to see the competition.
How do you take a concept that’s been around for 50 years or more and revive interest and get people as excited as you are?
First, they have to look at BNCT with an open mind and look at the advantages of the method. What could be achieved, if the proper things were done, in the proper order, with an open mind and the ability to look at types of cancer other than glioblastoma and with different compounds. This, of course, has not been the case, because of the previous concern only with the brain, and secondarily with melanoma.
Case in point, there are some very imaginative things that have been illustrated for internal organs, such as tumors of the liver. What I’m referring to here is work done in Pavia, Italy. Actually the results were announced a little over a year ago. What happened there was that the human patient, in this case a man with a very badly tumor-obstructed liver, was infused with a lot of BPA. When his blood level had reached a predetermined point of boron concentration, the liver was then excised ? taken out of his body — and kept alive as it was being irradiated outside the body. . The patient was kept alive by machines, then the liver was reinstalled and a few days later the patient went home.
That particular operation, the excorpus irradiation of body parts that can be sustained as they’re being radiated, is really neat. The patient, instead of trying to find a surrogate liver, provides his own transplant. This could work perhaps on other internal organs. The groups working on this in Pavia were a surgical group — again, surgeons led the way — and reactor people, engineers and physicists. The chemistry was BPA, which was available on the market commercially. That’s an imaginative application of BNCT.
The nice thing about BNCT is that you only have radiation, only have something cytotoxic, as long as the boron10 nuclei are exposed to neutrons. If you turn the neutrons off, the boron is just plain old boron and there’s nothing very toxic about that element. It’s like a light bulb: you can turn it off and on.
If you’re able to generate interest in this for a series of human trials and it is eventually approved by the FDA, how does it fit into current treatment infrastructure?
This would require a different type of radiation than what’s available today. Radiation oncology, as I know about it, is photon therapy, high energy X-rays. This radiation can be precisely delivered, targeted, and measured. The dosimetry is precise, the positioning of the radiation is precise, so it’s a nice way of doing it. But it has to be given repeatedly to a patient in many, many fractions from different directions. This is effective. But it requires many therapy sessions. Plus, it kills normal cells as well as tumor cells, though tumor cells are more susceptible to the radiation damage because they’re reproducing rapidly.
On the other hand, if you could get the BNCT selectivity feature working for you, you could slip in and target the tumor cells with boron, irradiate the whole area with a swarm of neutrons, and when the neutrons hit the boron 10 nuclei, you have very localized radiation of a very lethal sort. This is the production of particles that travel a very short distance. So it’s very localized. You do less collateral damage using neutrons than they would get in a similar dose delivered to tumor with traditional radiation. This dose can be delivered in perhaps one therapeutic session.
Is it something current radiation oncologists will have trouble picking up?
Not if they want to learn! One can expect a certain reticence to accept a new radiation method of this sort, to handle diseases which are currently handled using conventional radiation, since this could be considered as a competitor to existing methods.
If a person has established a clinic using conventional radiation, it’s natural that he would prefer to continue with this known course, rather than be diverted away by a new therapy which would require capital investment for equipment, plus learning and going into the unknown.
So there will probably be resistance of this sort. But I think that can be overcome if the results are truly beneficial to the patient. It can be overcome if the diseases to be treated are picked appropriately, so that you select those not readily curable by conventional photons for one reason or another. Also, if the knowledge required to practice this BNCT method is readily available, and if physicians want to learn it, they will do so.
What sort of timeline are you talking about and what happens if you can’t, at this juncture, get people turned on about this?
Here’s the situation: There have been clinical trials with glioblastoma multiforme at Brookhaven. There have been clinical trials in Japan. And some new studies have sprung up in Finland and Sweden. But always, it’s BPA being used against glioblastoma.
These clinical trials are basically repetitions of the Brookhaven trials that were just completed. Why this is being done is not clear to me. It’s like a movie of a horse race. The horse race occurs, a horse wins, and then you show the film over and over again, hoping perhaps that this time another horse will win. It’s not gonna happen.
So what do we do now? We look at people with melanoma like they’re doing at Harvard and MIT. But beyond that point, there seems to be no official push, if you will, to pursue this technology. The Department of Energy in the U.S. was the principal proponent, with money and nuclear reactors and all that good stuff. But the DOE has taken itself out of the business following the Brookhaven trial, which was viewed not as a failure exactly, but not as being successful to the point where they’d wish to invest more of their research money.
That leaves BNCT, at least on the clinical end, high and dry. To proceed, one must find a new way of raising money to finance new clinical trials against other diseases. This might not be easy, because the mindset in the U.S. government, at least, is that they’ve done enough for BNCT. Consequently, I believe the only immediately viable source of funding is from the public, people willing to donate money through a fund of some sort dedicated to the elucidation of the true nature or value of BNCT.
How much for the next chapter?
For a clinical lung cancer study, the chemistry and compounds leading up to that plus perhaps a second clinical trial for another disease such as head and neck tumors with liposomes, I think all that could be done for about $10 million. Of course, it would have to be a well-designed program.
But here’s the thing: For BNCT you need neutrons, you need boron chemicals, plus you need medical technology and dosimetry technology. All those things now exist, and you have the people still assembled and alive who are trained and knowledgeable in these areas. Given another three to five years, with attrition of people’s interest, people retiring, and so forth, plus the loss of availability of neutron sources, and it could combine to be disastrous to the field.
Put another way, if everything we know now crashes, 10 years from now if someone wanted to look at BNCT, and say, “Gee, this is a godsend the way neutrons and boron 10 do what they do, let’s go look at it in terms of medicine again,” to go back would cost many hundreds of millions of dollars.
We’re almost at the answer. We could find out more about the available scope of the therapy and find the one or more diseases that are much more common than glioblastoma multiforme that would hopefully be more amenable to BNCT therapy. But one has to move pretty quickly, because the people with the expertise are for one reason or another not going to be available.