This was an OK article, but clearly the author knows almost nothing about modern cancer diagnostics and therapy:
> Despite some early instances where genomic sequencing was performed on cancer patients, like Steve Jobs in 2009, the overall impact of the new science on healthcare has been modest at best.
Extensive genomic sequencing is at the heart of all translational and clinical center research, and attending a single cancer conference would disabuse the author of their misconceptions here.
Some treatments, such as those that target gene fusions, may still have been discovered at a pace several orders of magnitude slower at similar increases in cost. But we have hugely accelerated discovery through knowledge of a reference genome, whole transcriptome sequencing, and targeted sequencing of genes known to be associated with cancer (which is of course not all).
All the future developments in early detection, the "holy grail" of reducing cancer death, are based fully on large scale genomic sequencing.
Read through NCCN guidelines on cancer treatment, and all the newer diagnostics and therapies within the past 10-15 years are either directly genomic or greatly accelerated through the use of genomics in cancer.
> most cancer is still diagnosed and treated in a "classical" manner.
This is half-true at best.
As someone who's helped build FDA-approved genomic tests for cancer treatment, a huge proportion of cancers (in the US) are treated by sequencing and informing treatment based on sequence results. Especially among stages 3 and 4.
Diagnosis of early-stage (1 & 2) cancer via genomics is a difficult problem, because when the cancer is small, so is the amount of DNA it gives off. Liquid biopsy tests have a lot of trouble reliably distinguishing that small signal from noise, and it's not really cost-effective to test every single person for stage 3+ cancer via blood tests on a regular basis. Treatment of some early-stage cancers can suffer from the same problem, where it can be difficult to obtain a sufficient sample of cancer genetic material.
But the treatment landscape for late-stage cancers looks utterly different than it did 20 years ago. The idea that most stage 4 cancer is still treated in a "classical" manner in the US is patently false.
Genomics doesn't merely have massive potential, it's the backbone of research in cancer, even if the diagnostics are slimmed down to simple PCR tests or to even antibody staining on a slide, or to assigning drugs.
In some of the big cancers, sequencing is even a primary diagnostic:
* breast cancer: gene expression signatures are standard for determining the necessity of chemo
* colon: blood based sequencing for KRAS mutation is extremely common
* lung: matching kinase inhibitors to EGFR mutations is essential for therapy
* all solid tumors: tumor mutation burden is a companion diagnostic for the major immune checkpoint therapies that have revolutionized treatment over the past decade.
I know of almost no cancer that has not had massive changes in treatment, due to accelerated research possibly only through extensive use of genomics.
There’re many targeted oncology panels on the market, whole exome sequencing is getting more widespread in developed counties. These inform clinical decisions today, not in the future.
Thinking that being able to edit genes means you can control phenotypes is like thinking that poking a silicon die with a heated stick will let you display whatever you want on a screen. Technically, yes it will, but there are second-, third- and nth-order effects that are still extremely poorly understood.
Nor does the author understand the genetics of complex diseases. The “implications” claimed below are wrong because high heritable does not mean predictable across environments nor is there any genuine hope for selection of embryos for highly polygenic disorders.
> . . . prenatal methods will expand to highly heritable complex diseases: autism is 85% heritable, Alzheimer’s risk is 70% heritable, schizophrenia is over 75% heritable, and bipolar disorder is 70% heritable. The implication here is clear: we can eliminate many of these diseases through genetic screening of embryos or selective termination.”
There is a naive hope that a “polygenic risk” score of a condition can be predicted for complex diseases in the same way as for Huntington’s disease base in the number of CAG trinucleotide repeats. But polygenic risk scores are built on an assumption of additive and independent effects of DNA variants AND innumerable poorly defined environmental and developmental histories. This is just wishful thinking for most complex diseases. The concordance of diagnoses for schizophrenia among identical twins is no better than about 50%. That fact should give everyone pause.
8 comments
[ 2.5 ms ] story [ 36.5 ms ] thread> Despite some early instances where genomic sequencing was performed on cancer patients, like Steve Jobs in 2009, the overall impact of the new science on healthcare has been modest at best.
Extensive genomic sequencing is at the heart of all translational and clinical center research, and attending a single cancer conference would disabuse the author of their misconceptions here.
Some treatments, such as those that target gene fusions, may still have been discovered at a pace several orders of magnitude slower at similar increases in cost. But we have hugely accelerated discovery through knowledge of a reference genome, whole transcriptome sequencing, and targeted sequencing of genes known to be associated with cancer (which is of course not all).
All the future developments in early detection, the "holy grail" of reducing cancer death, are based fully on large scale genomic sequencing.
Read through NCCN guidelines on cancer treatment, and all the newer diagnostics and therapies within the past 10-15 years are either directly genomic or greatly accelerated through the use of genomics in cancer.
This is half-true at best.
As someone who's helped build FDA-approved genomic tests for cancer treatment, a huge proportion of cancers (in the US) are treated by sequencing and informing treatment based on sequence results. Especially among stages 3 and 4.
Diagnosis of early-stage (1 & 2) cancer via genomics is a difficult problem, because when the cancer is small, so is the amount of DNA it gives off. Liquid biopsy tests have a lot of trouble reliably distinguishing that small signal from noise, and it's not really cost-effective to test every single person for stage 3+ cancer via blood tests on a regular basis. Treatment of some early-stage cancers can suffer from the same problem, where it can be difficult to obtain a sufficient sample of cancer genetic material.
But the treatment landscape for late-stage cancers looks utterly different than it did 20 years ago. The idea that most stage 4 cancer is still treated in a "classical" manner in the US is patently false.
In some of the big cancers, sequencing is even a primary diagnostic:
* breast cancer: gene expression signatures are standard for determining the necessity of chemo
* colon: blood based sequencing for KRAS mutation is extremely common
* lung: matching kinase inhibitors to EGFR mutations is essential for therapy
* all solid tumors: tumor mutation burden is a companion diagnostic for the major immune checkpoint therapies that have revolutionized treatment over the past decade.
I know of almost no cancer that has not had massive changes in treatment, due to accelerated research possibly only through extensive use of genomics.
> . . . prenatal methods will expand to highly heritable complex diseases: autism is 85% heritable, Alzheimer’s risk is 70% heritable, schizophrenia is over 75% heritable, and bipolar disorder is 70% heritable. The implication here is clear: we can eliminate many of these diseases through genetic screening of embryos or selective termination.”
There is a naive hope that a “polygenic risk” score of a condition can be predicted for complex diseases in the same way as for Huntington’s disease base in the number of CAG trinucleotide repeats. But polygenic risk scores are built on an assumption of additive and independent effects of DNA variants AND innumerable poorly defined environmental and developmental histories. This is just wishful thinking for most complex diseases. The concordance of diagnoses for schizophrenia among identical twins is no better than about 50%. That fact should give everyone pause.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7465115/