It was probably at some point in my 20’s that I understood the concept of “you are what you eat”. Call me slow, but I finally figured out that eating burritos/pizza purchased from a gas station and heated in a hazmat microwave – for breakfast – was a bad idea.
Next I remember walking through a museum in London (it likely was something less exotic, like St. Louis) that described improvements to the body’s overall health at each time interval after one stops smoking: after 30 days your sense of smell improves; at 90 days your taste buds function better; at 3 years your circulation returns to normal; at 10 years your risk of death or slightly less than had you never started smoking. I’m clearly making the specifics up, but Dr. Google will fill in the missing blanks for those interested.
Now think about oil spills in Alaska and in the Gulf of Mexico, and all the little fish that feed the larger fish that are then served on our dinner plates at fancy restaurants across the country. Or at Micky D’s & Kantucky’s, more likely. One nugget of wisdom I learned recently is that dates, those sweet jewels from mother earth, attach to heavy metal ions as they pass through your intestines, ridding your body of chemicals such as mercury, for instance.
The point of all these anecdotes is that everyone reading this can make the logical link between input and output. You put bad things into your mouth, your body feels bad as a consequence (hello, hangover). You put bad info into a computer algorithm, wrong information comes out the back.
The predictive value of any oncology diagnostic is only as good as the technological sophistication of the methods being used. For anyone who has read previous posts, is interested in science, and wants to change the world, let’s now explore IHC & FISH…
IHC & FISH are the two standard of care methodologies used to determine if a woman diagnosed with breast cancer is a candidate for some of the most modern drugs on the market today, known as monoclonal HER2-targeting antibodies. The result of the IHC/FISH assays also gives the physician valuable information about the patient’s prognosis, ie whether he/she is facing an aggressive form cancer.
Seems like pretty valuable technology, if it were my family member who was being treated, I would certainly be interested to learn the outcome of this IHC/FISH result (tongue-in-cheek; I would demand that my doctor use more advanced technologies).
IHC is a relatively inexpensive technology, around 200 bucks, and is most commonly ordered for breast cancer tumors. To simplify the science, IHC stains the proteins found on the tumor cell and a pathologist looks under a microscope and makes a subjective assessment of whether he/she is seeing lots of HER2 proteins, or few HER2 proteins.
If he/she is unsure, the result is “equivocal”, or “beats me”. For anyone who read my previous posts, remember that Lab A will only get the same answer as Lab B ~80% of the time – for the same patient’s tumor drawn on the same day under the exact same circumstances.
Talk about garbage in, garbage out. It gets worse though, in case you weren’t incensed enough with just that level of subjectivity.
Not only is the “reproduceability” of the IHC assay only 80%, but inherent flaws of the assay also lead to improperly stained HER2 proteins in the first place. Finally, after a pathologist looks under a microscope he/she has the inauspicious task of determining whether this woman has aggressive breast cancer, and whether he/she will be denying access to potentially life-saving treatment. Sounds like wild guesswork to me. I’ll even take it one step further and call it: Cowboy Medicine.
So the pathologist doesn’t want to go out on a limb and he gives it a 2+ IHC score, which means “I don’t know” or scientifically “equivocal”. In 99% of all scenarios, this biopsy sample will then automatically trigger a FISH analysis. Let’s now dive into FISH…
(For those reading this far, anxious to know what these acronyms stand for here you go: ImmunoHistoChemistry & Flourescent In Situ Hybridization. Now aren’t you glad I didn’t mention that previously? Don’t even worry about what they mean, from now on let’s just say IHC is Incredibly-unHolyCrap and FISH is downright FISHY.)
As to FISH, let’s expose the fraud once-and-for-all: it is a ratio, plain and simple. It is a ratio of DNA. But wait a minute, aren’t we trying to find out if a HER2-targeting agent will work, and isn’t HER2 a protein? Why is FISH looking at DNA if HER2 is a protein? I hope at least a few of my readers don’t click away at this biology sarcasm, but the point is FISH is looking at a breast tumor’s DNA, and then makes a prediction of HER2 protein expression.
Remember: garbage in, garbage out. Predictions based on flawed assessments are totally flawed at the end of the day. Back to FISH, it gets worse…
So everyone now understands that FISH is a ratio of DNA, but in any ratio you need 2 items to compare, right? Good on the two of you keeping score!
So what is the denominator, if tumor DNA of predicted HER2 expression is the numerator? Here’s where I even lose myself in the science: the denominator is what’s known as a “centromere of chromosome 17”. But don’t even worry about it, we don’t even need to know what that means, does or says for now. Just trust me on this, you’ll be able to solve this riddle without knowing DNA from proteins from centromeres from s’mores crunch bars.
After the pathologist who said “I don’t know” to an IncredibyunHolyCrap result, another pathologist is looking into yet another microscope and is ready to make YET ANOTHER subjective call. I’m shocked, join the revolution.
In this subjective call, he/she is literally counting up little specs of fluorescent dots. It’s easy though, because they’re either red or green. The pathologist is counting up the red dots, and dividing by the number of green dots – and voila! Patient prognosis and drug options at your service.
But hold on a second, does FISH find every single DNA cell in the breast tumor? Not a chance, there would just be a big glow under the microscope in that scenario. FISH is looking at a subset of maybe 1% of the tumor, and making a prediction of the entire tumor. Fascinating, no? Or is that what we humans would describe as appalling?
So now what if the FISH pathologist counts 12 reds and 12 greens? That would be HER2 negative, since the ratio is a 12:12, also known as 1 for us math wiz’s. In this case the patient will not be given the drug Herceptin, since this patient is found to not have many HER2 proteins (remember though, FISH is looking at DNA, not proteins; quite the stretch of logic, but there it is).
So what if a different pathologist looks under the same microscope for the same patient at a lab directly across the street? Consider this scenario: he/she now counts 2000 reds and 1900 greens. “Holy canoly, that’s a lot of reds…but that’s also a lot of greens. Divide and carry the one…the ratio is 1..05”. In this scenario, the patient is also deemed to have a HER2-negative tumor, and the patient is denied access to potentially life-saving treatments.
Now I’m no math genius, but 2000 looks to be a lot more than 12 in my mind, so how can these two results provide the exact same clinical outcome? Could it be a function of garbage in, garbage out? For anyone keeping up, the red dots are DNA that are thought to code for HER2 expression. Lots of red dots tells me there is a lot of coding going on, so simply taking a ratio of DNA could lead to totally arbitrary results.
And what are those results? Apparently only 50% of patients respond to Herceptin, and the entire scientific/pharmaceutical community is wondering why and desperately trying to find bigger and better drugs to give those 50% of patients who don’t respond to Herceptin.
When I ask the question, why is the medical community still using IHC & FISH, the response is: that’s all there was available since we began treating with Herceptin, it’s all these docs know from their time at medical school; also, the pathology/laboratory/hospital community makes money from the interpretation of IHC/FISH results so there’s a reluctance to cutoff that revenue-stream; and for the most part, Herceptin seems to work; finally, for those patients for whom Herceptin doesn’t work, there’s another drug on the market called XYZ.
For those that are firmly aboard the OncoRevolution train, you’ll remember that the other drug available to HER2-positive tumors is a drug that crosses the blood-brain-barrier, whereas Herceptin does not.
Brain metastasis occurs much more frequently for patients on Herceptin vs. patients not on Herceptin – but don’t confuse cause and effect here. Herceptin doesn’t cause brain metastasis, it simply does not cross the blood-brain-barrier and thus has no ability to mitigate tumor development within the central nervous system.
That other drug does cross the blood-brain-barrier. If I was a doc, I would want to know if my patient has a ton of HER2 expression, or relatively low levels. The max score of IHC is 3+, and IHC is looking directly at HER2 protein expression. FISH is looking at DNA, and it’s just a ratio of DNA, not HER2 expression.
Come on folks, someone has to be keeping up with all this. Let’s change the world. Check that, let’s improve the way our families and neighbors are being treated by oncologists, pathologists, and the entire pharmaceutical industry. IHC is even being used in ongoing clinical trials for colorectal cancer, nonsmall cell lung cancer, ovarian cancer, etc. I promise there are better solutions out there than IncrediblyunHolyCrap.
Is it time for me to research what the heck a hashtag is? Help make this info go viral, join the OncoRevolution.