Cancer Prevention: Where does the new GRAIL test fit in?
Article by Robert Huizenga MD
Cancer isn’t inevitable like death and taxes - but it’s frighteningly common. A newly born child has about a 1 in 2 lifetime chance of being diagnosed with cancer. The good news is that scientists at the Harvard School of Public Health estimate that 50 to 75% of American cancer deaths can be prevented if the following Ten Cancer Commandments are consistently employed:
Get HPV and hepatitis A/B vaccinations (worldwide,
this is #1 most effective cancer prevention)No tobacco or secondhand smoke
Daily exercise
Optimal weight (measure waist size not weight)
Minimize sugar and insulin enhancing simple carbs, favor fish and poultry over red meat which may bump the risk of colon cancer and aggressive forms of prostate cancer, eat the rainbow (deep greens of spinach, blues of blueberries, whites of onions, and garlic, orange of carrots, reds of tomatoes, purples of eggplant, and so on) and avoid dietary carcinogens in processed lunch meat and bacon.
Limit daily alcoholic drinks to one (females) or two (men) to lower the risk of mouth, esophagus, liver, colorectal, and breast cancer.
Safe sex (HIV, HPV, hepatitis A/B/C and chlamydia all associated with cancer)
Avoid environmental toxins such as asbestos fibers, benzene, aromatic amines, and polychlorinated biphenyls (PCBs)
Avoid excessive sun and radiation (no convincing link between electromagnetic radiation or radiofrequency radiation and cancer)
Do not take supplements to decrease cancer risk (Vitamin D may be the exception, it may make cancer cells less aggressive and less apt or spread)
Adhering to these lifestyle commandments is especially important for persons with strong family histories of cancer or the minority of patients who possess one of 91 known genetic variants that
predispose to cancer (above our already high baseline risk). The problem is - let’s be honest - these Commandments are extremely difficult to adhere to. Only a tiny fraction of Americans – less than 5% - is able to meaningfully comply with all of the above.
And none of these commandments addresses the number one risk factor for cancer (and heart disease, strokes, Alzheimer’s, and Parkinson’s) – age! Cancer is predominantly an old age disease -
60-70% of cancers occur after 60 and if are fortunate enough to have it discovered at a curable stage, other cancers often can and do occur. I have several older patients who have had 4 unique potentially fatal
cancers. In the near future, the best cancer prevention – already possible in mice and currently being tested in humans - will be age reversal. (Human age reversal will be the subject of my next frontline report).
After doing the absolute best with standard prevention, we must rely on early
cancer detection.
Physical exams (enlarged lymph nodes, breast nodules, testicular lumps, or skin lesions)
Screening blood tests (leukemia, multiple myeloma, prostate cancer)
Screen urinalysis (bladder, renal cancer)
Radiographic studies* (chest, pancreatic, breast, brain cancers) (*remember one of ten cancer commandments is to avoid excessive ionizing radiation – one partial solution is to use MRI’s almost exclusively with the exception of ultra-low-dose chest CTs (lungs not well visualized by MRI’s) in high-risk former smokers)
Ultrasound (reproductive tract, thyroid cancers)
Upper and lower endoscopies with or without ultrasound (gastroenterology cancers)
Available prostate and lung liquid biopsies (to increase the odds that an elevated PSA or lung nodule is cancerous)
Once cancer spreads, the chance of cure or 5-year survival drops dramatically for all cancers. For example, for colon, breast, lung and pancreatic cancer, the 5-year survival is about 91/92/56/39% if diagnosed with only local disease (Stage I) vs 14/9/5/3% if the cancer is found to have spread distantly (Stage IV). Unfortunately, screening tools as currently used only inconsistently alert doctors to early cancer. Colon, breast, lung and pancreatic cancer are found early (Stage I) only 39/63/15/13% of the time.
The GRAIL test is a new machine learning technology aiming to improve on this dismal record of finding cancer in a treatable form. The GRAIL blood test based biopsy detects cell-free DNA (cfDNA) fragments from 50 types of cancer cells that spill out into our circulation.
The GRAIL test is resulted out as YES CANCER or NO CANCER (only a small % are reported as indeterminate). The test sensitivity (probability of having a positive test, should the person have cancer) is 55% and its specificity (the probability of having a negative test, should the person not have cancer) is 99%.
My typical patient over 50 in for a routine physical has about a 1% probability of having undiagnosed cancer (the “pre-test probability); let’s say a selected patient with increased risk factors might have a 2% probability of undiagnosed cancer. A GRAIL NEG result would indicate that the individual had a 99.5% chance of not having cancer. A GRAIL YES result indicates a 44% chance of having cancer (amazingly the test is able to identify
the origin of cancer in 90% of cases). A specific workup is of course initiated to confirm cancer. If no cancer is found, repeat studies are indicated in 3-6 months as in some cases the cancer was found so early that none of our current tests were able to localize the small collection of cancer cells.
One problem, however, is that at this time, cancerous cf DNA markers appear to be harder to detect in Stage I cancers compared to later stages of the disease. Indeed, for all cancers, GRAIL test sensitivity increases from 20% for stage I, to 45% for stage II, to 80% for stage III, to over 90% for stage IV cancers. Fortunately, for the more severe cancers like pancreatic, colon, esophagus, stomach - the chance of finding stage I cancer is much higher (40%). Hopefully, these numbers will improve with time as the test artificial intelligence feature uses the data from millions of new samples to further hone GRAIL’s detection of specific cfDNA cancer signatures.
Bottom line, for now, the GRAIL test is optional. It costs about $1000 and is not covered by insurance. It is a reasonable yearly test for those persons at increased risk of cancer - older, more overweight, less fit individuals, those with strong family histories of cancer or with positive cancer genes or smokers.
For these higher-risk patients, a negative test will be reassuring (>99.5% chance of no cancer).
A positive test will be more problematic – and anxiety-provoking! There will be about a 50% chance that the test was in error (false positive) and a 50% chance that previously undetected cancer is present (true positive). The final kicker is the “new cancer” might be so small it’s not detectable with any current screening tests for up to half a year!
I’m old, so I will take the test next week.
Last add! One more example (it is a long list) why all doctors need to seriously reconsider mindlessly entering their patient’s information on electronic “healthcare” systems that our government and local hospitals self-serving tout.
The 21st Century Cures Act, which includes the Information Blocking rule that took effect April 5, requires healthcare professionals to deliver to the patient's electronic portal the results of their tests and studies in 16 categories as soon as they are electronically available, regardless of whether the physician has had a chance to read those results or talk with the patient about them, put them in context, and discuss further options for care. The GRAIL test is not best released in this manner.