St. Jude scientists answer the question ‘when does cancer start?’

Things that scare us and hold power over us often appear formless, challenging to identify, and difficult to escape. However, there is a mantra that can be helpful when it comes to tackling fear:

“Name it, and you tame it.”

The Greek physician Hippocrates (460-370 B.C.) described a growth appearing from a mass of swollen blood vessels as “karkinos” – a crab. The long vessels spreading out from the growth reminded him of the legs of the crustacean.

“Name it, and you tame it.”

We have named this disease cancer. But we have not tamed it, at least not yet. We have made vast leaps and strides, but still…perhaps we have not named it properly. Perhaps we need a deeper understanding of what cancer is and how to describe it.

For that, scientists are going back to the beginning, the very start of disease: its developmental origins. By identifying not just the cause, but the moment at which cancer arises, we can begin to label it for what it really is. Cancer is not the embodiment of fear waiting in the dark. It’s a misalignment in a particular pathway. A mutation to a gene. A single cell that steps out of line. Researchers are now appreciating that the road to comprehensive and tailored treatment begins at the root of the disease – and to understand what is going wrong to cause cancer, to really name it, may be the best way to finally tame it.

At St. Jude Children’s Research Hospital, scientists in the Department of Developmental Neurobiology are studying childhood cancers through the lens of developmental origins of disease, building the necessary understanding to name, and thus classify, these tumors. Leading to breakthroughs in the way they are treated.

For example, take medulloblastoma. Medullo, referring to the medulla oblongata, the brainstem. And blastoma, a Greek word, meaning “bud” or in this context, “embryo”. A poignant name for the most common malignant brain tumor in children. Scientists at St. Jude, among others, proved that medulloblastoma comes in different subgroups. It can be defined by the WNT or SHH pathways, or it can be Group 3 or Group 4. Similarly, take retinoblastoma, named Retino, referring to the retina, a cancer that appears in the retina of young children. In fact, almost exclusively young children.

Where does this pediatric cancer stem from, what is its root cause, its origin? Scientists at St. Jude are working to find out.

“Name it, and you tame it.”

Cancer that hits early in life

Science has come a long way in describing and understanding medulloblastoma in children.

“[In the past] if a child was diagnosed with a brain tumor such as medulloblastoma, pathologists would look at morphology – what the tumor cells look like under the microscope,” explained Paul Northcott, PhD, St. Jude Department of Developmental Neurobiology. “They'd look for morphological features, organization and disorganization of cells to discriminate a neuronal tumor from a glial tumor from an ependymal tumor. In retrospect, it was somewhat primitive compared to where we are today.”

Northcott’s lab explores the diversity of medulloblastoma, a type of cancer which affects up to 500 people, mostly children, each year in the US. This stemmed from observations that even patients with the same diagnosis of medulloblastoma responded wildly differently to treatment.

“Some kids would respond to therapy and do fairly well, and then others were just completely unresponsive and resistant,” said Northcott. “Clearly, there was heterogeneity in the eyes of the clinicians from an early stage. They knew that there was going to be underlying biological differences that would explain why some children respond to therapy and others do not.”

To understand the origin of a disease like medulloblastoma, researchers need to untangle the complex processes involved in cell development, a lofty challenge. Few are more aware of this than Michael Dyer, PhD, St. Jude Department of Developmental Neurobiology chair.

“Not only is there incredible diversity in the types of cells in our body, but you have to make millions of identical copies of each cell type, all functioning exactly the same way,” Dyer explained. “We understand this is guided by changes in gene expression, which are incredibly dynamic, but then there's all sorts of additional environmental components. Every kind of cue imaginable plays into this incredible orchestra.”

Dyer’s lab studies the developmental origins of cancers with a focus on, among others, retinoblastoma. For Dyer, retinoblastoma represents the prototypical developmental tumor.

“We now understand that the cancer starts during fetal development and is diagnosed at birth or the first few years of life,” said Dyer. “But once you've made it past childhood, you're likely never going to get retinoblastoma. It has to start during development.” Retinoblastoma stems from mutations in a single gene, RB1. This gene helps regulate cell growth, and if it is inactivated, it can lead to uncontrolled cell growth and cancer.

The simple origins mask a cascade of complications. Chemotherapy at such a young age can have complications; and for some patients, removal of the eye is the best route to treatment. Unfortunately, for some patients, RB1 mutation is also a red flag for things to come.  

“The RB1 pathway is implicated in virtually every human cancer,” Dyer expounded. “Because they have this gene mutated, they're highly predisposed to other kinds of cancers later in life.” St. Jude is now using clinical genomics to identify those patients and provide them with genetic counseling and cancer monitoring to catch subsequent cancers early so they can be treated.

Moving beyond broad classifications to single-cell genomics

Even in cancers with a single root of development, the tendrils of effect extend beyond one pathway. Indeed, most cancers are tied to multiple pathways. This means that to get at the true origin and identify paths to treatment, scientists need to organize cancers based on their source. With medulloblastoma, Northcott identified four.

“Researchers started to allude to heterogeneity that, based on gene expression profiling, could define unique groups,” Northcott said. “The first groups to emerge were the Wnt and Sonic Hedgehog (SHH) groups.” These pathways are crucial to telling cells when and where to grow and divide. The Wnt pathway gives a signal for cell division, whereas the SHH pathway organizes the resulting cells properly within our body.

“That really paved the way for what eventually became the four consensus medulloblastoma groups in our community,” Northcott continued, referring to his 2011 Journal of Clinical Oncology paper.

Along with the two named groups, distinguished by their ties to the Wnt and SHH pathways respectively, the other two groups were simply named Group 3 and Group 4. These medulloblastomas encompass a wider range of genetic alterations, making it challenging to pinpoint a single defining trait for each subgroup. The classification of medullablastoma is vital to correct treatment assignment.

“We've got to separate out the Wnts, the SHHs, the 3s and the 4s,” Northcott said, “to catalog their genetics separately and start to think about their underlying biology.”

Indeed, advances in single-cell transcriptional analysis make it possible to not just look at the different types of medulloblastoma but also identify different types of tumor cells and cell “neighborhoods” within the growths. It is now understood that different sub-groups of medulloblastoma likely stem from distinct individual cells of origin. Their own lineage. This has allowed researchers like Northcott to unravel the complexity of the disease on a completely new level, with spatial and temporal aspects.

“The distinct transcriptional profile that eventually results in a full-blown tumor is because it hit a particular cell type at a particular time point during development,” Northcott said. “Looking at MRI data, the Wnt and SHH groups are actually located at different anatomical regions of the cerebellum. They're not just randomly appearing anywhere in the hindbrain, they're quite distant, in fact.”

What about Group 3 and Group 4? Northcott revealed in a 2022 Nature publication that these are differentiated based on their time of onset.

“They all occupy the fourth ventricle of the cerebellum, which suggested that they might be related in terms of their developmental biology,” Northcott said. “But in terms of age of onset, the Group 3 medulloblastomas tend to be diagnosed in infants to 16-year-olds, whereas Group 4 would be 5-year-olds to adults. So, they have slightly overlapping ages of onset, but enough of a difference that you can say, "Well, Group 3 looks like it's earlier. Group 4 looks like it's later."”

Given how our understanding of these classes has grown through this targeted research approach, Northcott is excited about the future of therapy.

“We're in an era now where, based on all of our retrospective analyses of clinical cohorts and all the biological insights we've gained,” he stated, “we know that there's certain groups and subtypes that are quite responsive to certain therapies and agents that we can tailor therapy for according to their molecular identity.”

Untangling the pathways of early cell development

The more we know about the origins of these diseases, the closer we get to a cure.

“The retinoblastoma gene was the first human tumor suppressor gene ever identified,” Dyer pointed out. “It was seen to be a gene that predisposes to this rare childhood cancer, but researchers initially thought it was unlikely to have broader impact across cancer." When its link to numerous other cancers was shown, Dyer acknowledged the paramount importance to getting to the bottom of retinal cell development and to understand the driving forces behind retinal tumor formation. This also emphasizes the importance of studying rare childhood cancers to advance our understanding of all human cancers.  

The route a stem cell takes to become a mature retinal cell passes it through stages of growth and differentiation, during which the cells are immature retinal cells called progenitor cells. In a pair of papers published in Neuron in 2017 and 2019 as part of the St. Jude-Washington University Pediatric Cancer Genome Project, Dyer and colleagues demonstrated that the point at which retinoblastoma kicks off is likely during this period of change from rapid growth to differentiation of progenitor cells.

To study this further, they successfully developed small 3D retinal organoids in the laboratory which accurately represent the disease. Published in Nature Communications in 2021, this advancement was paramount to fully exploring the developmental origins of retinoblastoma.

“The goal is to, first and foremost, save lives, and then second, save some vision,” Dyer said. “The treatments that have been developed here have gotten us to a point where even in the most challenging cases, they do a really good job of both.”

Dyer also utilizes the same single-cell techniques as Northcott to understand another cancer of particular intrigue, rhabdomyosarcoma (RMS). RMS affects areas of the body where skeletal muscles form during embryonic development.

“We can now map different RMS tumor cells to different developmental stages,” Dyer explained, reflecting on his 2022 Developmental Cell publication. “What we found is that despite the fact that it's a tumor, the RMS cells still remember their developmental origins, and try their best to go through the different developmental stages that a normal cell would.”

The implications for the cancer’s developmental origin in therapy are widespread. While initial treatments to patients with RMS are relatively well-received with a 70% survival rate, recurrence after treatment has a 5-year survival rate below 20%.

“When we treat a patient with chemotherapy, all those different cell stages of the cancer aren't equally susceptible,” Dyer said. “We're killing certain populations, but there's a subset that survive. Rather than some intrinsic drug resistance, we think this is more about the developmental stage of the tumor cell. It's a very different way of thinking about cancer and development.”

What’s in a name?

Developmental origins in cancer matter. As scientists dig further into how this multifaced disease originates, they gain more power over it and chase away the fear of the unknown. By identifying the pathway (or multiple pathways) and chains of effect that result in tumor growth, scientists can more accurately name cancer ever more specifically – not just medulloblastoma, but the WNT subgroup of medulloblastoma. This knowledge brings value that can change the way these diseases are treated, but it is a long road.

Cancer looms over all humanity, often appearing formless, challenging to identify and difficult to escape from. But researchers like Dyer, Northcott, and many others are building the knowledge necessary to truly identify, classify, understand and conquer this disease. To get to that end, they need to find its beginning.

“Name it, and you tame it.”