Working from the same playbook to characterize rare sarcoma

Tumors are complex collections of cells. The differences between individual cells within tumors impact treatment success. Understanding this process requires a large, multifaceted dataset collected from many patients. But for childhood cancers, many of which are rare diseases, it can be challenging to assemble a robust amount of data. 

Such is the case for rhabdomyosarcoma, a type of pediatric cancer that develops in soft tissue, such as muscles, but can occur anywhere in the body. Only about 350 new cases of rhabdomyosarcoma are diagnosed in children each year in the United States, which means developing a comprehensive understanding of the disease and how to treat it is an involved process. 

This is particularly true for patients whose disease recurs after treatment. Recurrence happens when cancer cells survive following treatment. Understanding cell-type diversity cuts right to this issue; by understanding those rare cells that remain after treatment, scientists can learn to make therapy more targeted and effective. 

“When I was a trainee, we used single-cell sequencing to generate maps of the different types of cells present in rhabdomyosarcoma,” explained Anand Patel, MD, PhD, St. Jude Departments of Oncology and Developmental Neurobiology. “This allowed us to pinpoint a specific cell type responsible for recurrence.” Patel and his mentor, Michael Dyer, PhD, Department of Developmental Neurobiology chair, presented their findings in a 2022 publication in Developmental Cell, highlighting the diversity of cell types in rhabdomyosarcoma cell populations.

At the same time, two other research groups presented similar findings, with differing nomenclature and classification strategies. Crucially, considering the rarity of rhabdomyosarcoma, all studies were limited in sample size, contributing to the differing viewpoints. There appeared to be only one solution for everyone to get on the same page: combine and compare. 

Joining forces and datasets to address rare cancers

“One of the things that’s amazing about science, and particularly pediatric cancer, was our immediate instinct to combine our datasets,” Patel said. Consequently, the three groups, including Patel at St. Jude; David Langenau, PhD, at Massachusetts General Hospital; and Beat Schäfer, PhD, and Marco Wachtel, PhD, at the University Children’s Hospital of Zurich, pooled all of their data together, now totaling 72 samples, and had another look. “For the community to move forward as a whole, we need to have set terminology, so everybody is speaking the same language,” expressed Patel.

This exercise in collaboration, published recently in Nature Communications, allowed the groups to hammer out a consensus on the classification of rhabdomyosarcoma cell populations. “We were able to establish the tools necessary for anybody anywhere in the world with single-cell or bulk RNA sequencing data to apply it to their data,” Patel said. Additionally, the study demonstrated the importance of strength in numbers when tackling rare diseases, such as rhabdomyosarcoma. “This is a very rare disease. But together, we were able to start to see things that each individual group couldn’t see.” 

There are two types of rhabdomyosarcomas: fusion-negative and fusion-positive. They are defined by whether the cancer is driven by cells harboring an abnormal gene created by fusing two different genes. While fusion-negative rhabdomyosarcoma is more common, accounting for about 70% of all cases, fusion-positive rhabdomyosarcoma patients have poorer outcomes. 

“It was an open question of what was going on in fusion-positive rhabdomyosarcoma during treatment,” Patel explained. “But because we were able to pool all this data, we had enough fusion-positive samples to identify a new cell subtype that nobody had reported before.”

The group’s comprehensive annotation of rhabdomyosarcoma cells revealed four major cell subpopulations: 

• Undifferentiated progenitor cells 
• Differentiated cells that resemble mature skeletal muscle cells 
• Proliferative cells that lack markers associated with a specific differentiated state 
• Inert ground state cells that do not show any cell state markers at all

The researchers also noted the grayness between these subpopulations, highlighted by the appearance of transitional cell states.

Larger dataset yields surprising findings

The analysis also showcased some surprising findings, notably a cell subtype that more closely resembled neuronal development. “This was very surprising because this is a tumor that for decades we’ve thought of as recapitulating muscle development,” Patel explained. “Something about the treatment of these particular tumor cells allows them to jump a guardrail and go into the neuronal lineage. It’s fascinating biology.” 

This curious finding sets the stage for further research into how this is happening and what impact such a transition has on outcomes and treatment. Preliminary analysis suggests that the transition to a neuronal state occurs because of treatment and is not permanent. 

“We think it’s still a rhabdomyosarcoma and that the cells transition during therapy and then go back — it’s a very plastic situation. We’re actively recapitulating this in the lab so that we can start to dissect this biology better,” Patel said. 

As Patel continues to explore the nuances of rhabdomyosarcoma cell differentiation, it’s clear that collaboration and sharing of resources will allow him and his colleagues to continue tackling the differences exhibited by fusion-positive rhabdomyosarcoma cases. “We came together purely with the idea that we just want to make sure that everybody is speaking the same language,” Patel explained. “Once we put all that data together and had this experience of working with each other, then it became obvious that we need to look deeper.”