Although research has produced advances in the detection and treatment of many adult cancers, survival rates for childhood malignancies, including neuroblastoma, have not improved significantly over the last 30 years.
Researchers at the Hospital for Sick Children in Toronto have been using an internally developed high-throughput assay to screen libraries of currently approved therapies in the hopes of identifying compounds that are effective against neuroblastoma-initiating cells, but that spare normal stem cells.
The investigators screened more than 5,000 compounds and identified 47 that were potentially effective against neuroblastoma. Among those 47 compounds was rapamycin, made by Wyeth, on which the researchers are currently focusing.
Kristen Smith, a postdoctoral fellow at the Hospital for Sick Children, spoke with CBA News this week about the investigators’ work, which was presented at the American Association for Cancer Research-National Cancer Institute-European Organization for Research and Treatment of Cancer International Conference on Molecular Targets and Cancer Therapeutics, held last week in San Francisco.
How does the assay work?
The high-throughput assay involves screening two cell sources in parallel. We used neuroblastoma tumor-initiating cells isolated from an advanced stage, multiple-relapse neuroblastoma patient as well as normal pediatric stem cells isolated from human skin. The ability to compare tumor-initiating cells with normal stem cells from a similar developmental origin allows us to identify compounds that are selectively toxic against the tumor cells while sparing normal tissue.
We plated 3,000 dissociated cells per well in a 96-well plate — we also developed the screen for 384-well plates — [then] treated with compound immediately, and scored cell survival using Alamar Blue at approximately 54 hours. The cells were plated in serum-free media supplemented with bFGF, EGF, and B27. We screened three compound libraries: LOPAC, Spectrum, and Prestwick.
We focused on these libraries, which are composed of approved drugs and some natural products, so that we would be able to rapidly translate any hits from the lab to the clinic. The libraries were screened at 5 µM, 1 µM, or 0.4 µM, and hits were confirmed in duplicate.
IC50 experiments were performed for compounds that were more toxic against the neuroblastoma TICs. Forty-seven compounds showed dose-dependent selective toxicity against the neuroblastoma TIC and were selected for secondary screens.
We identified 40 novel compounds for neuroblastoma treatment as well as seven that have been or are currently used for neuroblastoma treatment. The compounds fell into several classes including DNA-damaging agents, digoxin-family members, anti-malarial agents, and protein-specific antagonists.
Secondary screening involved a sphere-formation assay. We wanted to determine whether the selected compounds targeted tumor-initiating cells or bulk tumor cells so we used sphere formation as a surrogate assay of TIC survival. Following these assays, we have begun in vivo testing using a neuroblastoma xenograft mouse model.
Can you further discuss your animal model?
We use two different mouse models: one for preliminary work and one for future studies. Our first model is a xenograft model. We take the neuroblastoma TICs that we isolated and we put them into the inguinal fat pad of non-Scid mice and we let the tumor grow to about .5 cm2, and then we started treatment. We are testing these drugs with established tumors, so it is a pretty stringent test. The reason that we do it this way is that is very easy to measure the tumor if we do not have to kill the mouse to follow the course of the disease or anything like that.
We treat with whatever compound for approximately two weeks, until the control animals or the vehicle-treated animals are sick and they have to be sacrificed. At that point we sacrifice the drug-treated animals and run all of our follow-up experiments.
The second model is an orthotopic model. In that case we inject the TICs into the adrenal fat pad, which sits right above the kidney. The reason we go to that site is it is where the majority of neuromuscular primary tumors begin. These animals develop metastases similar to those of human patients, so it lets us test a lot of different things in one animal.
Do you have a timeline for the next phase of your project?
We are right in the middle of the animal testing, so we have gotten rapamycin into our animal testing and we are doing some more biochemistry in the lab to try and figure out what is going on and what the difference is between the SKPs and the neuroblastoma cells. Why are the neuroblastoma cells so affected by this drug while the SKPs are not really affected at all?
We have another five compounds that we are moving into the animal model and we are working on two of those right now. The work is on a small scale, so it is not going that fast. As we get the results from those, then we will start to see if we can use them in patients. We do have a compassionate use program at Sick Kids.
If we get compounds that pass the animal test, as a last ditch attempt we are able to try some of these other things. They have either failed all of their therapies and are ineligible for another trial or there is no trial that they haven’t tried. We are able to get this stuff into the clinic very quickly. Right now, though, we are just waiting on the animal models.
Are you planning to publish any of this work?
We would like to publish it and will probably get something pulled together within the next few months. Again we are waiting on the animal stuff to see what will be the most important thing to focus on.
Will the team be focusing on any of the other 47 hits that have been
Which of these 47 hits looks to be the most promising?
Right now, we are focusing on some of the anti-malarial compounds because of their good safety profile in children as well as two specific protein antagonists. I'd prefer not to mention the names directly because we do get quite a bit of attention from patient groups.
Until we've gotten positive results from the animal models, we don't mention the exact compounds. I hope you understand that we don't want to unnecessarily raise the hopes of parents.
You mentioned that you are testing these cells against other compounds. I assume you are testing neuroblastoma TICs and SKPs?
Yes. The TIC isolation and culture is work that is currently in press at Cancer Research. What we were able to do was take either tumor samples or bone marrow samples from patients, because we see all neuroblastoma patients that come through Ontario.
We were able to grow cells from these samples such that when you inject them back into animals, they form a tumor that looks just a neuroblastoma. It takes only 10 cells to form these tumors, and they fulfill most of the criteria for a cancer stem cell or a tumor-initiating cell.
The use of TICs is the really novel thing about this work because very few people are able to grow them. What is particularly interesting is that they may represent the cells that cause the relapse of the disease, which is the real failure of therapy right now. You may be able to treat the tumor and it goes away for a little bit, but it always comes back.
By targeting these cells specifically, we hope we are actually getting the cells that cause the relapse. What we have been able to do is get cells from several patients.
Although the initial screen was done on a sample from one patient, we later went back and confirmed several of these hits with samples from multiple patients. It is interesting because most of the drugs will hit most of the patient samples. A few compounds demonstrate some variability between patients, though. That is interesting to us from the perspective of trying to design patient-specific therapies.
We are also interested in kind of a general therapy, however, so it’s really nice that we are getting compounds from these screens that are hitting multiple patient samples. Maybe they are something that we could use generally.
You mean against other types of cancers?
Yes, that is another interesting thing. Among the 47 hits that we came up with were compounds that were shown to target other TICs, such as leukemic stem cells, breast cancer stem cells, and glioblastoma.
So we are getting compounds that hit multiple things, for example, rapamycin, which has been shown to affect leukemic stem cells. There are some interesting overlaps. We have not explored them fully yet, but its something that we are also very interested in doing.
Why is your team focusing on rapamycin?
Rapamycin has great activity against neuroblastoma TICs from multiple patients. It has been shown to be effective in other cancer models and has good bioavailability. In addition, several rapamycin analogs are in clinical trials for solid tumors. Because of its good safety profile there is also interest in using rapamycin in combination with current neuroblastoma therapies.
One more important point: This work has been a team project. The screening assay was designed and performed in collaboration with Alessandro Datti, Jeffrey Wrana, and Jim Dennis at the Mt. Sinai/Samuel Lunenfeld SMART Facilty. The animal model is a collaboration with Libo Zhang, Herman Yeger, and Sylvain Baruchel at The Hospital for Sick Children. We have been funded by multiple groups including the Canadian Institutes of Health Research, National Cancer Institute of Canada, Lilah's Fund, the James Birrell Neuroblastoma Research Fund, the McLaughlin Centre for Molecular Medicine, and the Stem Cell Network.