Not to take anything away from the breakthrough but another line of treatment showed success too:
> One further patient, identified only as Drew, has been tumour-free for over four years after receiving immunotherapy.
The key part of the article, in my opinion, is this one (referring to the 13 year old):
> The scientific teams at Gustave Roussy cancer centre in France are now working intensively to understand what made his response possible. Their current hypothesis centres on what Dr Grill described as an extremely rare mutation in Lucas's tumour cells that made them unusually sensitive to everolimus by creating a dependency on the mTOR pathway. Researchers have begun creating tumour organoids, three-dimensional laboratory-grown models that mirror each patient's tumour's genetic characteristics, to test whether Lucas's specific cellular profile can be replicated.
Why is a randomized controlled study needed for these scenarios in a clinical trial. If the patient is known to have a fatal disease in the study, why not give them the proposed treatment? The alternative is much worse.
Because in most cases people die anyway. Let's hope this guy lives until he is 100 years old, but sadly there it is common that cancer is not completely eradicated and it may come back in 10 years or 20 or 5 or 1.
(Also, there are weird unexplained cures from time to time, so there is always a chance. Nothing is completely binary.)
So the question is if this new drug is useful. Let's say you give the new drug to 100 kids and the average life time is 3 years and 2 months. Is it good or bad?
You can compare it with other 100 kids with other treatments, but there are a lot of know unknowns and unknown unknown. For example the easy cases may revive the old drug because it works most of the time, and the new drug is only for the hard cases. If someone has kidney or liver damage, they may be excluded from the use of the new drug. The kids in the new drug may get a direct hot line to access to the hospital staff, that may improve their survival. The hospital may be in a rich/poor part of the city and the kids may get better/worse food than the average.
There are just too many variables that are impossible to control. So the solution is to give to half of the children the new drug and to the other half the old drug. So you ensure all the other variables are similar.
Perhaps the average of the kids with the other drug is 5 years and 1 months, and they got killed 2 years earlier. It's a risk with new untested drugs.
Perhaps the average of the kids with the other drug is 1 years and 3 months, and congratulations, they they doubled their expected life time. After a few more trials the new drug may become the standard treatment.
Are you talking about the word “randomly” in the following text?
> …the BIOMEDE study, which began in 2014 and compared three drug candidates: erlotinib, everolimus, and dasatinib. … Lucas was randomly assigned to receive everolimus.
Not to take anything away from the breakthrough but another line of treatment showed success too:
> One further patient, identified only as Drew, has been tumour-free for over four years after receiving immunotherapy.
The key part of the article, in my opinion, is this one (referring to the 13 year old):
> The scientific teams at Gustave Roussy cancer centre in France are now working intensively to understand what made his response possible. Their current hypothesis centres on what Dr Grill described as an extremely rare mutation in Lucas's tumour cells that made them unusually sensitive to everolimus by creating a dependency on the mTOR pathway. Researchers have begun creating tumour organoids, three-dimensional laboratory-grown models that mirror each patient's tumour's genetic characteristics, to test whether Lucas's specific cellular profile can be replicated.
Why is a randomized controlled study needed for these scenarios in a clinical trial. If the patient is known to have a fatal disease in the study, why not give them the proposed treatment? The alternative is much worse.
Because in most cases people die anyway. Let's hope this guy lives until he is 100 years old, but sadly there it is common that cancer is not completely eradicated and it may come back in 10 years or 20 or 5 or 1.
(Also, there are weird unexplained cures from time to time, so there is always a chance. Nothing is completely binary.)
So the question is if this new drug is useful. Let's say you give the new drug to 100 kids and the average life time is 3 years and 2 months. Is it good or bad?
You can compare it with other 100 kids with other treatments, but there are a lot of know unknowns and unknown unknown. For example the easy cases may revive the old drug because it works most of the time, and the new drug is only for the hard cases. If someone has kidney or liver damage, they may be excluded from the use of the new drug. The kids in the new drug may get a direct hot line to access to the hospital staff, that may improve their survival. The hospital may be in a rich/poor part of the city and the kids may get better/worse food than the average.
There are just too many variables that are impossible to control. So the solution is to give to half of the children the new drug and to the other half the old drug. So you ensure all the other variables are similar.
Perhaps the average of the kids with the other drug is 5 years and 1 months, and they got killed 2 years earlier. It's a risk with new untested drugs.
Perhaps the average of the kids with the other drug is 1 years and 3 months, and congratulations, they they doubled their expected life time. After a few more trials the new drug may become the standard treatment.
Are you talking about the word “randomly” in the following text?
> …the BIOMEDE study, which began in 2014 and compared three drug candidates: erlotinib, everolimus, and dasatinib. … Lucas was randomly assigned to receive everolimus.
[dead]