For the first time, researchers have identified at least three unique subtypes of a rare type of bone cancer, potentially transforming clinical trials and patient care.
A University
of East Anglia-led research project has been able to use advanced mathematical
modelling and machine learning called "Latent Process Decomposition"
to categorise patients with osteosarcoma into different subgroups using their
genetic data. Previously, all patients would be grouped together and treated
using the same protocols, which has very mixed outcomes.
While
genetic sequencing has previously helped to uncover different subtypes of other
cancers, like breast or skin cancer, for which those patients then receive
targeted treatment personalised to their cancer subtype, it has been much
harder to do this with osteosarcoma - a cancer that starts in the bone and
typically affects children and teenagers.
Lead author
Dr Darrell Green, of UEA's Norwich Medical School, said: "Since the 1970s
osteosarcoma has been treated using untargeted chemotherapy and surgery, which
sometimes results in limb amputation as well as the severe and lifelong side
effects of the chemotherapy.
"Multiple
international clinical trials investigating new drugs in osteosarcoma have been
deemed to have 'failed' over the last 50 plus years.
"This
new research found that in each of these 'failed' trials, there was a small
response rate (around five to 10 per cent) to the new drug, suggesting the
existence of osteosarcoma subtypes that did respond to the new treatment.
"The
new medicines were not a total 'failure' as was concluded; rather, the drugs
were not successful for every patient with osteosarcoma but could have become a
new treatment for select patient groups.
"We
hope that in the future, grouping patients using this new algorithm will mean
successful outcomes at clinical trial, for the first time in over half a
century.
"When
patients can be treated using targeted drugs specific to their cancer subtype,
this will facilitate a move away from standard chemotherapy."
The search
for kinder, more targeted treatments for osteosarcoma is an important area of
focus for Children with Cancer UK.
In 2021,
funding was awarded by the leading childhood cancer charity to the team at UEA
to investigate innovative ways to treat osteosarcoma.
Dr Sultana
Choudhry, Head of Research at Children with Cancer UK, said: "Investing in
pioneering research programmes is integral to driving forward our vision of a
world where every child and young person survives cancer.
"We
invest our fundraising into science because
we've seen how research can make a significant difference in the survival
chances of every child.
"By
funding groundbreaking research, we are not only advancing scientific knowledge
but finding gentler, more effective treatments for our youngest and most
vulnerable cancer patients.
"Our
hope is that the outcomes of this research project will improve the diagnosis,
treatment and long-term care for young cancer patients."
The survival
rate for osteosarcoma, a type of bone cancer, has stagnated around 50pc for the
past 45 years. This is mainly because the different subtypes of osteosarcoma
are not yet fully understood, as well as how the immune system around the
tumour affects it, or what causes the cancer to resist treatment or spread to
other parts of the body.
Scientists
are yet to identify the key biological markers that could help predict a
patient's outlook or how they will respond to treatment. These gaps in
knowledge are preventing progress in improving survival rates.
Previously,
researchers have tried to predict different types of osteosarcoma by using
certain computer methods, which suggests that there are distinct subtypes of
the cancer.
While this
was an important step forward, this doesn't fully account for the fact that
each osteosarcoma tumour can be very different from one part to another.
These models
also assume that each tumour can be neatly placed into one specific group, even
though tumours are usually made up of many different kinds of cancer cells.
This
variation within a tumour makes it harder to accurately predict how the cancer
behaves or responds to treatment.
In this
study, researchers used a more advanced method called Latent Process
Decomposition (LPD), which takes into account the differences within individual
tumours.
Unlike
earlier methods, LPD looks at the tumour as a mix of hidden patterns in gene
activity. These hidden patterns represent different "functional
states" of the tumour, and each state has its own specific gene expression
pattern.
The LPD
method figures out how many of these patterns are needed to describe a
particular tumour.
The research
uncovered three osteosarcoma disease subtypes, one of which was found to
respond poorly when treated with the standard chemotherapy drug combination
called MAP.
By grouping
patients based on these patterns, doctors could make more informed decisions
about treatment.
Researchers
acknowledged that key limitations of the study include a small dataset for the
LPD model development, and the incomplete clinical data in the validation
cohort.
Access to
tissue and linked clinical data is particularly challenging for osteosarcoma
due to the rarity of cases, limited biopsy material and the extensive
chemotherapy-related damage present in post-treatment samples.
Despite
these challenges, the LPD method proved to be reliable, as it identified
consistent subgroups of osteosarcoma across four different sets of independent
data.
Like any
machine learning tool, the results get better as more data is added.
No comments:
Post a Comment