“Chronic Myeloid Leukaemia (CML) is an aggressive type of blood cancer that requires life-long therapy. However, in most low and middle-income countries, long-lasting treatment with specialised drugs presents several challenges that impact the equality, accessibility and effectiveness of treatment for patients,” said Göran Karlsson of the Lund Stem Cell Center at Lund University. “Thus, improvements of current therapy regimens or new curable drugs would have an important impact for CML care in Africa.”
Although there has been significant progress in drug development and longer survival rates for CML there are still problems including access to high-priced drugs, side-effects, the need for extensive clinical care and the fact that the available treatments are not effective for everyone leading to treatment failures with associated risks of disease progression. Karlsson also explained that bone-marrow transplantations can be used when people fail drugs. “It does work and can be a cure but the result dependent on finding a genetic match.”
“However, large-scale analysis of the molecular content of individual cells (‘single-cell genomic analysis’) holds great promise for improving personalised treatment and even curing CML by identifying rare, persistent and relapse-initiating cancer stem cells, as well as by effectively identifying prognostic markers and new drug targets. However, even if an abundance of such single-cell data is being produced, comprehensive conclusions and consensus of these efforts are lacking.”
Karlsson, who is also a Wallenberg Academy Fellow, summarised the knowledge acquired by his group and others from single-cell research on CML stem cells over the last decade and showed how this research is coming closer to clinical reality.
He started by explaining the different types of stem cells – namely totipotent, pluripotent and multipotent – which together can generate all the cells of the body.
“Blood stem cells reside in the bone marrow and recreate themselves while producing all the functional cells of blood – red and white cells and platelets,” he explained. “Cancer stem cells can also recreate themselves through self-renewal and can mature to all the cells of the tumour.”
In CML there is a mutation in stem cells – which is always the same – a break in chromosomes 9 and 22. Because this break point is in the middle of two genes you get an abnormal fusion gene. This mutated gene is a dysregulated tyrosine kinase which makes cells super active. Lots of cells are produced but they are not functional, normal cells. This means there are, for example, less red cells transporting oxygen to the body and less immune-system cells to prevent infections.
Until the early 2000s survival rates with CML were low, but this changed with a major breakthrough in the early 2000s with the introduction of Imatinib (or Gleevec) developed by Novartis. Now the five-year survival rate is about 90%.
Karlsson explained that Imatinib was able to save the healthy cells but kill off the cancer cells and showed very positive results in phase I to III clinical trials. “In clinical trials 95% of patients who got it did well”. In 2001 the FDA granted approval for Imatinib in less than three months which is very unusual and the drug made the cover of Time Magazine in May 2001 – as a model example of molecular targeting of cancer.”
“But five years later there were trials where Imatinib was withdrawn in those patients where Imatinib had been most effective and no leukaemia cells could be detected. Within six months 60% of this stop patients had relapsed. So, they had not been cured by the Imatinib treatment and leukaemic stem cells were still there in their bone marrow, causing the relapses.”
“This meant that Imatinib was basically a lifelong treatment and very expensive,” he continued. “Even though patents for Imatinib have now expired, the 40% of patients who fail Imatinib treatment have to move on to another type of tyrosine kinase inhibitors which are still very expensive. And patients need regular monitoring and usually experience some side-effects. There is still a lot of research left to be done in CML”
Looking for a needle in a haystack
It had become clear that to understand the relapse and learn how to cure or develop better treatments for CML, it was necessary to be able to separate the leukaemic stem cells from the healthy stem cells of the bone marrow.
“But they are a needle in a haystack,” said Karlsson. “Healthy stem cells are only about 0.01% of bone marrow, leukaemic stem cells in a patient receiving treatment even less.”
And this is where single-cell genomics – which Karlsson was working on at the time – entered the picture.
He explained that all cells express a genomic programme (or molecular signature) that reflects the function of that cell and can be used as its fingerprint. “All cells express a range of cell surface proteins (or markers). If these markers correlate to a molecular signature, they can be used to isolate and analyse certain cell types.”
“If you find cell surface proteins that correlate with a molecular signature, you can isolate that particular cell type and study it,” he added.
He compared single-cell analysis vs. bulk analysis with a fruit salad analogy. “Bulk analysis is like making a a smoothie of the fruit salad where we have to guess which fruit is included. Single-cell analysis is like looking at each fruit separately,” he said. “Single-cell genomics describes each cell type’s constituents so you get a complete picture of all the fruits making up the salad, or the cell heterogeneity of a sample or tissue. It can identify rare cells and define which genes drive each cell type.”
By taking CML cells at diagnosis and then later during treatment they hoped to be able to see if you could predict the patient’s long-term response to drug therapies at diagnosis.
In comparing the distribution of cell types per patient, they found that those who responded well had lots of stem cells that had started to mature into platelets or red blood cells, while those who responded poorly had more inactive immature stem cells. “We therefore concluded that a poor response to Imatinib is associated with a high proportion of immature stem cells. These inactive/immature stem cells survive Imatinib but the stem cells with platelet/red blood cell signatures are Imatinib sensitive. Even more important was the amount of healthy stem cells that were present in the bone marrow in these patients. It seems that a high number of healthy stem cells was critical to get a good response from treatment. The response to Imatinib can therefore be diagnosed depending on the ratio of each type and the amount and quality of the healthy stem cells at diagnosis.”
“This means you can do testing and analysis upfront and recommend the correct treatment already at the day of diagnosis. If patients have a lot of immature/inactive cells you can give them stronger treatment options right away.”
Although it’s still early days single-cell analysis holds the promise of improving personalised treatment and even curing CML by identifying these rare but persistent cancer cells, as well as prognostic markers and drug targets.
“Hopefully we can also learn more about stem cells in general to know which mutations and molecular features that are important for treatment-response in other Leukaemias,” said Karlsson. “This work is also transferrable to other malignant blood diseases.”
Democratising data
But even if lots of single-cell data is being produced not everyone in the world can access and analyse it. Karlsson’s STIAS project is therefore trying to democratise single-cell genomics and make data available to all by generating a CML single-cell genomics atlas.
They have developed the Nygen Analytics analysis tools and database which is free for academics, doesn’t require a code or expensive hardware by reducing memory usage to that of a laptop, is cloud based and already contains hundreds of publicly available datasets.
“But it still needs CML data,” said Karlsson. “We are therefore taking single-cell data from the 10 most prolific studies to analyse and merge. This will include data on about one million cells (including clinical data) – free for all to use. The hope is that any expert in CML from doctors to researchers should have access to this valuable data independent of single-cell expertise or resources.”
They have also created an AI analysis tool – CyteType – a cell-characterisation package able to extract data, score it for confidence, compare it to existing knowledge and integrate it into a report. “CyteType is made to analyse huge datasets in a short time using all published knowledge in minutes or hours. It will find information in the data that would take years for us to come across.”
“It would be inspirational to cure CML,” he concluded. “We should be able to fix it. But along the way there is lots of method development and clinical trials – all of which is very valuable for other health-research areas.”