CRM1 Emerges as a Potential Target in Osteosarcoma in Novel Drug Discovery Model

Alexander Leandros Lazarides, MD, a resident at Duke University Medical Center

Alexander Leandros Lazarides, MD

In a high-throughput screen, investigators demonstrated similar in vitro and in vivo antitumor activity with CRM1 inhibitors in dog and human osteosarcoma—derived cell lines and xenografts, respectively, validating a comparative cross-species precision medicine pipeline that has the potential to enhance drug discovery in a difficult-to-treat disease.

In the study, whole-exome sequencing was performed on dog and human osteosarcoma models to identify potential driver mutations, according to lead study author, Alexander Leandros Lazarides, MD.

Among 2100 drugs that were screened, CRM1 inhibitors were found to be effective in killing dog and human cell lines in vitro and showed a significant reduction in tumor growth in these xenografts in vivo.

“The next steps for this research are to translate [these findings] into a clinical trial in canines,” said Lazarides. “We're striving for [a] precision medicine [method], but nothing has become standardized. That’s what we’re working toward with our mouse-dog-human pipeline. We're trying to [develop personalized approaches] that we can translate into [the clinical setting].”

In an interview with OncLive^®, Lazarides, a resident at Duke University Medical Center, discussed the need for better drug discovery models in osteosarcoma and provided insight into a new personalized medicine pipeline that has the potential to redefine drug development in the field.

OncLive: Could you discuss the unmet need in osteosarcoma treatment?

Lazarides: Osteosarcoma is a challenging cancer; it's one that has bimodal distribution, but disproportionately affects children. We made tremendous advances in the 1970s with the advent of chemotherapy, which led to a dramatic improvement in survival. Since then, we haven't made [too much progress].

There's a need for novel drug discovery, we just haven't found a way to do that and there are many reasons why. Our existing models are not sufficient for a heterogeneous disease like osteosarcoma. Osteosarcoma has many different mutations, so it’s challenging to target 1 pathway. For us, it’s a matter of finding a better model [that can help us discover] new drugs to treat our patients.

We’ve looked to canines to try and use a comparative oncology approach. There are many benefits to using canines. For one, they share the same environment as us. They also have a higher incidence of osteosarcoma, and their disease course is much more accelerated. What might take 5 years to study in a human will take just 1 to 2 years in a dog.

What was the goal of your study?

We wanted to validate our mouse, dog, and human pipeline, as well as the CRM1 pathway as a possible target in osteosarcoma. We used a 3-tiered approach in which we started with the patients, either canine or human, and collected tissue samples from them. Then, [we created] cell lines that [served as] our in vitro line. We had patient-derived xenografts, which we grew in mice, and [that served as our] in vivo testing line. We also ran genomic analysis on these tissues. Then, we brought all that back together [in the hope of] translating it rapidly into the clinical realm.

What were the results of this trial?

For this to be a functional approach, we had to prove that canines and humans respond the same way and have similar targets. That was the first aim of our study. Therefore, we compared the response of canines in human osteosarcoma cell lines with different drugs. First, we did that with a more limited drug screen. The drugs [we tested] are in use for several other diseases and cancers. We did a cluster analysis to identify which lineage it came from: canine or human. We had 5 humans and 4 dogs, and we were able to show that you couldn't cluster it; [we couldn’t determine whether the response to those drugs came from] dogs or humans.

We [expanded this analysis] to 2100 drugs. In a similar manner, using advanced statistics, we were able to show percent response to multiple drugs in these different cell lines, and show that there's a similar trend in responses between dogs and humans. That showed us that we can use canines and humans for a comparative oncologic approach. We then looked at CRM1 [inhibitors], which were one of the [drug classes] that popped up in our drug screen.

In our in vitro cell line analysis, we saw a very high percentage killing of cancer cells [with that drug]. It’s an interesting target because multiple drugs that targeted this pathway were found to be effective; we saw over a 90% killing of these cancer cells. We investigated it further and found that [CRM1] is ubiquitously expressed across several different cells and cancer cell types; it’s not specific to one cell line, which is important. We used the database out of Europe, which has genomic analysis, and found that when [a patient has] higher CRM1 expression, [they] have a worse outcome than those with lower CRM1 expression, further validating its use as a potential target.

CRM1 is an exportin nuclear export receptor. [CRM1] shuttles products from the nucleus into the cytoplasm of the cell, so they can be processed and turned into more functional proteins. In cancer, things are being upregulated and constantly produced at a higher rate. If we shut that down, [CRM1] can't produce all of the proteins and whatnot that it needs to survive. Then, we did an in vivo clinical trial in mice. We developed a bunch of different tumors in the mice and then tested [the drug]. We found that [the agent] has a very high effect on limiting the rate of tumor growth in mice compared with controls and other drugs. This [finding] kind of validated the model as a whole, as well as the CRM1 pathway as a target.

What are some of the commonly used drugs in osteosarcoma?

Several agents are available. Doxorubicin, cisplatin, and methotrexate are commonly used. Some other variations exist in that regard. We looked at these agents in our drug screen, and while methotrexate and doxorubicin had reasonable tumor-killing effect, we found that cisplatin did not [have the same effect] in these osteosarcoma cell lines.

This plays to the fact that [osteosarcoma] is a very heterogeneous disease that is very good at adapting and evading these different drugs. We really need to be thoughtful about the way we're treating the disease; we have to do this in a way that our traditional treatment approaches are not for all of our patients. We need to be more personalized with our approaches.

How might these data impact the field?

In the past 30 years, we haven't made any strides in our ability to treat patients, particularly those with metastatic disease. That's the big issue [that we face]. While patients who have nonmetastatic disease may have [a] survival [rate that is] upwards of 70%, those who have metastatic disease are looking at a 20% or 30% survival [rate]. We may be able to delay tumor progression in some patients, but we may not actually be able to cure them.

We need a way to identify drugs in a personalized manner that can be more rapidly translated back into the clinical realm to impact patient care. That's what we're trying to do with this study. By taking patients’ tissue and testing it in the lab, [we may be able to] determine a personalized medicine approach [that can be used to more effectively treat these patients]. A comparative oncologic approach comparing dog and human responses in a kind of paired manner has never really been done before; it's a novel idea for discovering new drugs.

What is the key takeaway from this research?

The most important takeaway is that this pipeline can work. The next steps are to translate [this research] into the clinical realm. We've been able to collect patient data and run tests on it in the lab. The next step is translational. Our study has shown that we can identify novel targets in a personalized manner.

Lazarides AL, Somarelli JA, Yang Z, et al. A cross-species personalized medicine pipeline identifies the CRM1 export pathway as a potentially novel treatment for osteosarcoma. Presented at: Musculoskeletal Tumor Society Annual Meeting; October 2-4, 2019; Portland, OR. Abstract 32. bit.ly/2uhhSQf.