Molecular Pathology: The wave of the future?

[futureresident]

---

Members do not see this ad.

Some say that molecular pathology is the greatest thing since sliced bread and will largely replace what we do on a regular basis; others more on the cynical side, say that we have a LONG way to go because of standardization issues.

I personally feel that while it has it's limitations, it's a powerful tool that has tremendous diagnostic, predictive, and prognostic utility. It already has an important role in hemepath and microbiology.

An attending of mine eloquently stated:

Right now, we have a blunderbuss approach to treatment using highly toxic chemotherapy, but with molecular pathology we could refine it to hit at precisely the Achilles heel of the tumor, at the one mutation which has given it the drive to grow.

What are your thoughts?

 

[LADoc00]

Some say that molecular pathology is the greatest thing since sliced bread and will largely replace what we do on a regular basis; others more on the cynical side, say that we have a LONG way to go because of standardization issues.

I personally feel that while it has it's limitations, it's a powerful tool that has tremendous diagnostic, predictive, and prognostic utility. It already has an important role in hemepath and microbiology.

An attending of mine eloquently stated:

Right now, we have a blunderbuss approach to treatment using highly toxic chemotherapy, but with molecular pathology we could refine it to hit at precisely the Achilles heel of the tumor, at the one mutation which has given it the drive to grow.

What are your thoughts?

Long ago I had this idea that molecular pathology would absolutely revolutionize pathology...changing the very nature of the field itself from clumsy subjective morphometry unchanged since the 1800s to a shiny, new molecular classification of all human disease. That idea was wrong for lots of reasons I couldnt possibly imagine at the time, but now I know better.

 

[SaltySqueegee]

From a different perspective. It's not that molecular diagnostics has failed completely, it is that it has yet to come of age.

You hit it on the head with the Standardization statement. Many labs are trying to create these systems, and characterize disease states according to these systems. But it takes years to do. This is not an over night process. A single (or small family of) biomarker(s) could take 2-3 years to develop. Many families of genes involved in pathological states have yet to be discovered still.

From my PI, he said that from his perspective, the FDA is finally coming to the point where they realize that non-standardized Micro-arrays are practically useless when it comes to treatment. How do you know the treatments or diagnosis/prognosis is what it is if you lack internal standards, interexperimental standards, and intergenic expression standards. Thus the FDA is now pushing the development of QA/QC projects for this. But this implementation will not be soon.

You may start to see the age of Molecular diagnostics in 15-20years. It will be gradual. Scientists are still developing the tools.

3rd year MD/PhD

 

[LADoc00]

From a different perspective. It's not that molecular diagnostics has failed completely, it is that it has yet to come of age.

You hit it on the head with the Standardization statement. Many labs are trying to create these systems, and characterize disease states according to these systems. But it takes years to do. This is not an over night process. A single (or small family of) biomarker(s) could take 2-3 years to develop. Many families of genes involved in pathological states have yet to be discovered still.

From my PI, he said that from his perspective, the FDA is finally coming to the point where they realize that non-standardized Micro-arrays are practically useless when it comes to treatment. How do you know the treatments or diagnosis/prognosis is what it is if you lack internal standards, interexperimental standards, and intergenic expression standards. Thus the FDA is now pushing the development of QA/QC projects for this. But this implementation will not be soon.

You may start to see the age of Molecular diagnostics in 15-20years. It will be gradual. Scientists are still developing the tools.

3rd year MD/PhD

Our knowledge of the complex molecular interactions that underlie the morphologic classifications of disease is infantile at best, dangerously incorrect at worst.

As a great California writer once said:
Its not what we know we dont know that's the problem.
Its what we think we know for sure, but is wrong that's the problem.

 

[PathOne]

Molecular diagnostics have made real and clinically significant contributions in very narrow fields, notably infectious diseases and heme. However, it's still essentially clinically useless when it comes to solid tumors, despite more than ten years attempts at "cracking the code" and billions and billions spend on developing diagnostic markers.

Still, the hope is still living and strong that it'll someday be possible to improve cancer diagnostics by molecular approaches. In reality, I strongly feel that we should work to make it happen, but be willing to accept that molecular markers will mainly be used to screen for patients who may benefit from very specific treatments (the HER2/Neu-story being a prime example). I simply don't see molecular pathology replacing conventional light-microscopy-and-immunstaining procedures anytime soon.
And it cannot yet be ruled out, that molecular pathology may end up like electron microscopy, i.e. basically a dead end, accept for a very limited number of pathology cases.

The truth is, that light microscopy for many years will remain the gold standards, and that the development of diagnostic markers in pathology will remain troubled by the fact that there's really, really few board certified pathologists out there that really understand molecular biology. The vast majority of researchers in the field are molecular biologists, who wouldn't recognize a tumor if it was hitting them in the face. Thus, there's tons of research which comes to a dead end due to the GIGA-principle (Garbage In, Garbage Out), because they're examining something which just isn't what they think it is... Quite sad, really.

 

[futureresident]

I agree that light microscopy will remain the gold standard for many years to come, but one cannot refute the role of molecular diagnostics as an ancillary tool to our standard practice today and in the future.

It's a matter of time before more vital observations will be made, standardization of techniqes will be achieved, and we will be armed with a whole new array of sophisticated techniques to add to our diagnostic arsenal. Over time, these techniques will be honed. Perhaps it may be sooner than 15-20 years. Neverthelsess, I feel it is our responsibility to contribute to this field as it allows for optimal patient care. and incoroporate it into our daily practice.

For heme, it's applications are very clinically significant, and also for micro. I would like to see growth in solid tumor diagnostics.

Is it possible that molecular diagnostics will have the same/simislar role in our practice as imunohistochemsitry?

 

[LADoc00]

I also want to add that 99% of basic science lab geeks in this field have no clue as to what would really benefit pathologists or aid in diagnosis. PIs are too concerned with chasing tenure or grant renewal. Nope folks, the only people with their eye on the prize sortof speak are the MBA types, who in chasing the almightly dollar actually come up with some good ideas from time to time. Which brings me to my next point, if you really want to advance science get an MBA, not a PhD.

 

[b&ierstiefel]

I agree that light microscopy will remain the gold standard for many years to come, but one cannot refute the role of molecular diagnostics as an ancillary tool to our standard practice today and in the future.

It's a matter of time before more vital observations will be made, standardization of techniqes will be achieved, and we will be armed with a whole new array of sophisticated techniques to add to our diagnostic arsenal. Over time, these techniques will be honed. Perhaps it may be sooner than 15-20 years. Neverthelsess, I feel it is our responsibility to contribute to this field as it allows for optimal patient care. and incoroporate it into our daily practice.

For heme, it's applications are very clinically significant, and also for micro. I would like to see growth in solid tumor diagnostics.

Is it possible that molecular diagnostics will have the same/simislar role in our practice as imunohistochemsitry?

I agree with some of the above posts that molecular diagnostics has made an impact in some fields like heme but the utility for general surg path is is an "infantile" state now. Who knows, maybe it'll get better. I would envision that it would aid in delineating subcategories of different types of tumors and aid in further improvements in tumor taxonomy. The frustration with subjectivity and interobserver variability with respect to many tumors is real. I don't think molecular will do anything to upend morphology anytime soon.

 

[b&ierstiefel]

I also want to add that 99% of basic science lab geeks in this field have no clue as to what would really benefit pathologists or aid in diagnosis.

Haha...so true. Quite a few experiments I've seen in papers lead to conclusions that are quite physiologically implausible in light of some of the pathology I've learned this year. Research to aid in diagnosis would be quite useful, I agree. Unfortunately, they don't lead to high profile papers which is what you really need to advance yourself in basic science...it's all about publish or perish.

 

[Matte Kudesai]

I also want to add that 99% of basic science lab geeks in this field have no clue as to what would really benefit pathologists or aid in diagnosis. PIs are too concerned with chasing tenure or grant renewal. Nope folks, the only people with their eye on the prize sortof speak are the MBA types, who in chasing the almightly dollar actually come up with some good ideas from time to time. Which brings me to my next point, if you really want to advance science get an MBA, not a PhD.

Have you luddites seen companies like this one?

http://www.historx.com/

 

[PathOne]

I agree that light microscopy will remain the gold standard for many years to come, but one cannot refute the role of molecular diagnostics as an ancillary tool to our standard practice today and in the future.

It's a matter of time before more vital observations will be made, standardization of techniqes will be achieved, and we will be armed with a whole new array of sophisticated techniques to add to our diagnostic arsenal. Over time, these techniques will be honed. Perhaps it may be sooner than 15-20 years. Neverthelsess, I feel it is our responsibility to contribute to this field as it allows for optimal patient care. and incoroporate it into our daily practice.

For heme, it's applications are very clinically significant, and also for micro. I would like to see growth in solid tumor diagnostics.

Is it possible that molecular diagnostics will have the same/simislar role in our practice as imunohistochemsitry?

Unfortunately, it's not as easy as it may seem. The theory behind molecular diagnostics is beautiful, but the reality is somewhat more complicated. Essentially, the past ten years of research in molecular diagnostics of solid tumors has been a wasteland. We simply cannot develop reliable and reproductible standards for molecular diagnostics, and often, the cost and time spent doing it is simply prohibitive compared to the tried-and-true ways. As one of the greatest dermatopathologists once dryly remarked: "Phil LeBoit and others doing research in molecular pathology run around the country saying "we're almost there", and they've been saying that for years"....

To give an example: One of the problems in using traditional staining is, that you're really staining for the cells, not the cancer itself. So if you stain nevi with S-100, it'll stain the same for a benign lesion and a malignant one. However, with the aid of the Mark I Eyeball of a Dermpath, the morphology will tell you what's what.

Now, do molecular diagnostics on the same tissue. Look for something we know pretty well, like BRAF mutations. They're present in most malignant lesions. Great. However, they're equally, and sometime more so, present in perfectly benign lesions. Why? We simply don't know. And we don't have adequate tools to look at what you might call morphology at the molecular level.

So that's one of the reasons why you're seeing tons of high-impact papers on tumor biology, describing a never-ending symphony of cellular pathways. But whenever you're seeing experiments that attempt to develop clinically relevant diagnostics, they either produce great results on very few samples (often 20 or so), or they produce results that simply aren't statistically significant.

With time and energy, I am sure that good things will emerge from this current black hole. Especially techniques like FISH and CISH are beginning to hold true promise. But it's a lot tougher than anyone ever imagined...

 

[Matte Kudesai]

Have you luddites seen companies like this one?

http://www.historx.com/

One more time

 

[PathOne]

Have you luddites seen companies like this one?

http://www.historx.com/

Have never heard of it, but then again, another biotech company selling products that can't be used in clinical diagnostics...

From their website:
"The AQUA™ platform applies to basic and clinical research as well as drug discovery, including pharmacogenomics and pharmacodiagnostics."

Read: No, we're not FDA approved either....

 

[Matte Kudesai]

Have never heard of it, but then again, another biotech company selling products that can't be used in clinical diagnostics...

From their website:
"The AQUA™ platform applies to basic and clinical research as well as drug discovery, including pharmacogenomics and pharmacodiagnostics."

Read: No, we're not FDA approved either....

Wait a few years chief. Just pointing out that there are some bright people working on some amazing concepts.

As mentioned before amazing ideas such as flow cytometry became a diagnostic reality in heme that we cannot imagine living without.

The same thing will happen with diagnosing solid tumors.
There will be FDA approval.
And it will change diagnositc pathology witin 15 years.
The use of FISH TCRbeta rearrangments, array CGH, are only a few technologies that will impact pathology in the very near future.

 

[PathOne]

As I previously wrote: We're almost there. Only problem is, that that's been said for years now. In Feb 2000, J. Craig Venter and Francis Collins published the data on the complete human genome. At the time, it was heralded as a seismic event, which would quickly lead to better diagnostics and treatment for a wide veriety of diseases. Frankly, little has happened since.

Yes, Flow Cytometry has its uses, although limited in scope. Same with electron microscopy, and, currently, PCR, Northern and Western Blotting, CISH and FISH (to name the most widely used). Yes, it's also possible to get vast amounts of genetic information by using stuff like microarrays and SNP-analysis. Only problem is, that it's not the universal panacea that many people thought it was, and still think it is.
Personally, I think that molecular diagnostic will prove to be helpful in specific instances. But they'll remain auxillary to light microscopy for the next couple of decades.

 

[Matte Kudesai]

Unfortunately, it's not as easy as it may seem. The theory behind molecular diagnostics is beautiful, but the reality is somewhat more complicated. Essentially, the past ten years of research in molecular diagnostics of solid tumors has been a wasteland. We simply cannot develop reliable and reproductible standards for molecular diagnostics, and often, the cost and time spent doing it is simply prohibitive compared to the tried-and-true ways. As one of the greatest dermatopathologists once dryly remarked: "Phil LeBoit and others doing research in molecular pathology run around the country saying "we're almost there", and they've been saying that for years"....

To give an example: One of the problems in using traditional staining is, that you're really staining for the cells, not the cancer itself. So if you stain nevi with S-100, it'll stain the same for a benign lesion and a malignant one. However, with the aid of the Mark I Eyeball of a Dermpath, the morphology will tell you what's what.

Now, do molecular diagnostics on the same tissue. Look for something we know pretty well, like BRAF mutations. They're present in most malignant lesions. Great. However, they're equally, and sometime more so, present in perfectly benign lesions. Why? We simply don't know. And we don't have adequate tools to look at what you might call morphology at the molecular level.

So that's one of the reasons why you're seeing tons of high-impact papers on tumor biology, describing a never-ending symphony of cellular pathways. But whenever you're seeing experiments that attempt to develop clinically relevant diagnostics, they either produce great results on very few samples (often 20 or so), or they produce results that simply aren't statistically significant.

With time and energy, I am sure that good things will emerge from this current black hole. Especially techniques like FISH and CISH are beginning to hold true promise. But it's a lot tougher than anyone ever imagined...

As you well know the interobserver variability in regard to melanocytic lesions is quite high (benign vs malignant). Although the eyeball is an amazing tool, it is the owner of the eyeball, their reputation, derm political affiliation (Ackerman, Leboit, Crowson, McNutt etc) that aids in the generation of a DX.
Ultimately the patients clinical course post diagnosis that dictates whether the initial DX was correct. I guess this is why dermatopathologists are the most sued pathologists. We need more powerful tools to aid the eyeball and make things a little less qualitative.

 

[PathOne]

As you well know the interobserver variability in regard to melanocytic lesions is quite high (benign vs malignant). Although the eyeball is an amazing tool, it is the owner of the eyeball, their reputation, derm political affiliation (Ackerman, Leboit, Crowson, McNutt etc) that aids in the generation of a DX.
Ultimately the patients clinical course post diagnosis that dictates whether the initial DX was correct. I guess this is why dermatopathologists are the most sued pathologists. We need more powerful tools to aid the eyeball and make things a little less qualitative.

Yes, pathologists, even the most experienced ones, makes mistakes and use guesswork. But show me a single instance where you can use molecular diagnostics alone in making a diagnosis. After all, we've had the ability to make PCR reactions for more than 20 years now, and in that period a staggering amount of money has been poured into making it happen. It might happen someday, but so far, we just haven't seen it, so I really would be extremely careful about setting a timeframe for the death of traditional pathology...

 

[Matte Kudesai]

Flow Cytometry has its uses, although limited in scope. .

How can you say that the use of flow is limited in scope. Can you imagine signing out difficult heme cases without having flow??? 😕

 

[PathOne]

Flow Cytometry works, because it's designed and used for a specific set of diseases (in heme). The promise of molecular diagnostic pathology was that it could be used for precision diagnostics of all cancers. That simply isn't the case, and in fact, molecular pathology has been utterly unable to move beyond heme and into solid tumors in clinical cancer diagnostics.

 

[Matte Kudesai]

Yes, pathologists, even the most experienced ones, makes mistakes and use guesswork. But show me a single instance where you can use molecular diagnostics alone in making a diagnosis. After all, we've had the ability to make PCR reactions for more than 20 years now, and in that period a staggering amount of money has been poured into making it happen. It might happen someday, but so far, we just haven't seen it, so I really would be extremely careful about setting a timeframe for the death of traditional pathology...

Traditional pathology will never die.... If by traditional pathology you mean using a microscope. We just have to be careful not to polarize ourselves in either direction. I find it disheartening when molecular pathologists diss traditional pathologists or vice versa. IHC that is routine now developed over twenty years as well. Anyway it is a cool time to be in this profession. Lots of amazing things going on.

 

[pathstudent]

I also want to add that 99% of basic science lab geeks in this field have no clue as to what would really benefit pathologists or aid in diagnosis. PIs are too concerned with chasing tenure or grant renewal. Nope folks, the only people with their eye on the prize sortof speak are the MBA types, who in chasing the almightly dollar actually come up with some good ideas from time to time. Which brings me to my next point, if you really want to advance science get an MBA, not a PhD.

Interesting....I met this ready to graduate undergrad a few days ago. He was a double major molecular biology/biochemsitry. I assumed he was another future MD. He said that was his plan when he started but now is doing an MBA/PhD program as "MDs are the future teachers, fireman, and policeman of the world in that they will deserve a lot more pay for what they do." I sort of nervously gulped and wished him luck.

 

[1Path]

Anyway it is a cool time to be in this profession. Lots of amazing things going on.

Absolutely! I saw some amazing research while in Path dept at the NIH which easily combined molecular (such as CISH and FISH) with traditianal pathology. In fact, it's hard for me to imagine the field of pathology (based on this expereicne and what I've seen at USCAP) as consisting of one and not the other.

 

[PathOne]

Amazing, yes, but mostly not really relevant clinically. I doubt that molecular diagnostics can really be used in staging, but there ought to be some breaks in the future when it comes to sub-classification. Mostly, though, we're all still crossing our fingers and hope that it'll work one day.

 

[SaltySqueegee]

I think the statement above about the human genome completion being a seismic event (2000) put it best. Read: it was not that long ago. Hence, people are still trying to put the puzzle together to see how it works, after that the intricate cellular interactions will (hopefully) become more understandable. With each gene having literally hundreds of unique transcription factors and upstream mediators which have potential for creating cellular malignancy and dedifferentiation, we still have a way to go to find the real big players. But it will (hopefully) happen soon enough.

So, yeah, as was said above, "it's an exciting field, and will be for some time to come."

Here is to Molecular Biology 👍

 

[SaltySqueegee]

I think the statement above about the human genome completion being a seismic event (2000) put it best. Read: it was not that long ago. Hence, people are still trying to put the puzzle together to see how it works, after that the intricate cellular interactions will (hopefully) become more understandable. With each gene having literally hundreds of unique transcription factors and upstream mediators which have potential for creating cellular malignancy and dedifferentiation, we still have a way to go to find the real big players. But it will (hopefully) happen soon enough.

So, yeah, as was said above, "it's an exciting field, and will be for some time to come."

Here is to Molecular Biology 👍

Okay... so I've had an additional year in molecular biology research. I originally quoted the era of molecular genetics in medicine to occur in 15-20 years. I now forecast my vote for 25-30 years. Give or take 20 years... ...

... 🙄

 

[PathJet]

i think molecular is already integrated into many large centers as a part of treatment decisons- so whether we think molecular will be important or not for DX might be immaterial if private companies are justifying there existance and publishing articles in the NEJM about the utility of their product-which may be excellent im not sure i dont administer chemotherapy.

. ONe area which this has changed markedly is breast cancer treatment in oncotype DX.

so initially my thoughts are --well didnt at some point someone thought EM would be the wave of the future---------and look how that turned out--

but molecular is pretty much here to stay we just need to decide the integration of it into our practice or it will be integrated into other practices with or without us etc...

attached is a 2004 article

 

[docbiohazard]

Molecular definitely needs to be integrated with pathology as the research comes along... as a research oriented student, thats one of the things that attracted to me to pathology, the opportunity for growth in translational research.

I also think the EM comparison, well, EM is EM. It's a great technique for ultrastructure but the sample handling and processing is cumbersome, and the equipment is pretty large, expensive and the like. I tend to believe molecular technology, with appropriate standards, will follow a curve of becoming smaller and cheaper, and will be easily integrated into the clinical laboratories of many major hospitals...

BH

 

[SaltySqueegee]

Okay... so I've had an additional year in molecular biology research. I originally quoted the era of molecular genetics in medicine to occur in 15-20 years. I now forecast my vote for 25-30 years. Give or take 20 years... ...

... 🙄

I'll be a bit more specific. The role of molecular genetics in complex (not Mendelian) disease susceptibility traits. Most cancers, autoimmune disorders, cardiovascular disease, etc. susceptibility. Those diseases which environment are the predominant factor, but genetics can tip the scale in either direction.

 

[DarksideAllstar]

i think molecular is already integrated into many large centers as a part of treatment decisons- so whether we think molecular will be important or not for DX might be immaterial if private companies are justifying there existance and publishing articles in the NEJM about the utility of their product-which may be excellent im not sure i dont administer chemotherapy.

. ONe area which this has changed markedly is breast cancer treatment in oncotype DX.

so initially my thoughts are --well didnt at some point someone thought EM would be the wave of the future---------and look how that turned out--

but molecular is pretty much here to stay we just need to decide the integration of it into our practice or it will be integrated into other practices with or without us etc...

attached is a 2004 article

Without going into too much detail, I've heard first-hand criticism of that article by several faculty members at a couple of different institutions, including one where a fellow gave a talk essentially dissecting their methodology.

 

[Neddy]

Have you luddites seen companies like this one?

http://www.historx.com/

The founder of that company happens to be one of my favorite pathologist-scientists - he's faculty at my former med school, and I was able to talk to him a few times about his research and plans. He's under absolutely no illusion that quantitative immunohistochemistry will replace morphology to any significant degree. The goal is to allow pathologists to get different kinds of information that we can't get now with the H&E, not (for the most part) to do the same thing we already do in a more standardized (and likely far more expensive) way. No one's come up with a molecular tool at any price that can handle one thousandth of the diagnostic subtleties that a good pathologist with a good light microscope can, but if you really want precision on whether that Her2 is 2+ or 3+, yeah it will probably be a computer doing that in a few years. No big loss IMHO.

 

[Neddy]

i think molecular is already integrated into many large centers as a part of treatment decisons- so whether we think molecular will be important or not for DX might be immaterial if private companies are justifying there existance and publishing articles in the NEJM about the utility of their product-which may be excellent im not sure i dont administer chemotherapy.

. ONe area which this has changed markedly is breast cancer treatment in oncotype DX.

so initially my thoughts are --well didnt at some point someone thought EM would be the wave of the future---------and look how that turned out--

but molecular is pretty much here to stay we just need to decide the integration of it into our practice or it will be integrated into other practices with or without us etc...

attached is a 2004 article

My experience is that Oncotype Dx hasn't really caught on in big academic centers. Most of the academic med oncs I've spoken with will use it in very selected circumstances, once they've considered all the traditional prognostic factors and either they or their patient are still on the fence about chemo. It's not cheap, and docs tend to be suspicious of its validity for anything but run of the mill IDC NOS. The test is very redundant of more traditional modalities (ER, PR, HER-2, and morphologic grade), and it's not actually clear from their publications that it adds very much to a methodical consideration of those factors plus stage.

I think they're doing better in the private practice world, where different incentives are in play, and frankly, the oncologists are more likely to know their pathologists through written reports alone, rather than as part of an interdisciplinary team.

 

[PathJet]

My experience is that Oncotype Dx hasn't really caught on in big academic centers.....

I think they're doing better in the private practice world, where different incentives are in play, and frankly, the oncologists are more likely to know their pathologists through written reports alone, rather than as part of an interdisciplinary team.

i am a little slow in general as a human being so im not sure what you are saying mainly re
.....who is doing better in the private practice world? pathologsists, oncologists, private companies

just to clarifiy my point ...

i dont necessarily support these types of assays or necessarily think they are clinically valuable- for example the result of the "genomic health" is an integer score --can you imagine dictating a score as a final dx WTF???--

"LEFT PAROTID, FINE NEEDLE ASPIRATION:
- 64 (see comment)"

Comment: 64 is associated with a lesional neoplasm. 32% of the 20,000 patients studied by this method had an associated neoplasm. Clinical correlation is advised."...

i mean this is basically the type of result you get with of one of these tests.

however this is a company which is motivated to involve itself into clinical carecare like other companies...

Genomic Health Inc. (NASDAQ: GHDX)
Open: 20.62
Mkt Cap: 568.98M
High: 20.70
52Wk High: 24.68
F P/E: -299.00 Yield:
N/A Low: 20.08
52Wk Low:
12.98
Shares: 28.11M
Vol: 19,125.00
Avg Vol: 132,000.00

this is only one company.

whether we all want to sit around ponder how molecular pathology is going to influence the future is sort of a moot point it is influencing us now and in the future to come-- so we might as well like learn as much about it and become as entrenched in the field as we can

 

[PathJet]

dup

 

[NKcell]

Molecular diagnostics have made real and clinically significant contributions in very narrow fields, notably infectious diseases and heme. However, it's still essentially clinically useless when it comes to solid tumors, despite more than ten years attempts at "cracking the code" and billions and billions spend on developing diagnostic markers.

Still, the hope is still living and strong that it'll someday be possible to improve cancer diagnostics by molecular approaches. In reality, I strongly feel that we should work to make it happen, but be willing to accept that molecular markers will mainly be used to screen for patients who may benefit from very specific treatments (the HER2/Neu-story being a prime example). I simply don't see molecular pathology replacing conventional light-microscopy-and-immunstaining procedures anytime soon.
And it cannot yet be ruled out, that molecular pathology may end up like electron microscopy, i.e. basically a dead end, accept for a very limited number of pathology cases.

The truth is, that light microscopy for many years will remain the gold standards, and that the development of diagnostic markers in pathology will remain troubled by the fact that there's really, really few board certified pathologists out there that really understand molecular biology. The vast majority of researchers in the field are molecular biologists, who wouldn't recognize a tumor if it was hitting them in the face. Thus, there's tons of research which comes to a dead end due to the GIGA-principle (Garbage In, Garbage Out), because they're examining something which just isn't what they think it is... Quite sad, really.

here's my thoughts on the future (however distant that may be):

molecular pathology will largely replace pathology as we know it, and to make matters worse, radiologists in the form of molecular imaging, will largely replace pathologists (in-vivo > in-vitro diagnosis)

bear in mind, i have absolutely no evidence to support my idea, except to say that the head of the nih is a radiologist, and a lot of research $ is begining to head into that field

 

[PathJet]

very interesting.....

i heard a very learned pathologists predict the fusion of pathology with radiology into a specialty called diagnostic medicine,


(his heavily visited blog)
http://feeds.feedburner.com/LabSoftNews

i guess the surgical specimen volume is heavily weighted towards biopsies in many organ systems

does any one know the percent/number decline in gastric resections since the discovery/ elucidation of H. pylori gastritis?

 

[Neddy]

here's my thoughts on the future (however distant that may be):

molecular pathology will largely replace pathology as we know it, and to make matters worse, radiologists in the form of molecular imaging, will largely replace pathologists (in-vivo > in-vitro diagnosis)

bear in mind, i have absolutely no evidence to support my idea, except to say that the head of the nih is a radiologist, and a lot of research $ is begining to head into that field

Keep in mind, though, that most "molecular imaging" techniques require the administration of a novel chemical agent to a patient, and the safety and regulatory hurdles one has to overcome to make something like that available for general use are staggering in comparison to tissue-based diagnostic techniques. How many truly novel imaging agents/modalities have been approved by the FDA in the last 10 years? I'm no expert, but I'll bet you could count them on one hand (actually, I can't think of any). The basic technology behind FDG-based PET scanning has been around since the 70's, and thirty years later there's still relatively limited availability.

That said, I agree that pathologists need to keep pushing the traditional boundaries of the field to keep our turf from shrinking. Molecular path, and maybe even molecular imaging will be part of the picture, but I'd argue that traditional pathology skills will be the mainstay much longer than some people think. Radiology will probably make the physical exam obsolete (if it hasn't already) before it drives the H&E to extinction.

 

[PathJet]

NICE!!!

that is what i like to hear....

also if you ever do QA with autopsies at your institution it seems one of the most frequent discrepancies is with imaging findings --that and missing an occult fungal infection .... dont you think?

 

[CameronFrye]

The fusion of pathology and radiology was discussed in a recent CAP Today article. I think UTSW is doing some of the best work with it. Anyway, I thought the most interesting part was a quote by one of the radiologists in the research group. He basically said that pathologists don't think outside of the box and don't work to develop new technologies, whereas it's kind of ingrained in radiologists.

The article made it sound like radiology would capture these images, but they would probably still need someone with path knowledge to interpret them.

 

[Neddy]

Anyway, I thought the most interesting part was a quote by one of the radiologists in the research group. He basically said that pathologists don't think outside of the box and don't work to develop new technologies, whereas it's kind of ingrained in radiologists.

I think that's a valid criticism of pathology, although to be fair, pathologists are limited in their ability to take the lead in crossover research by their lack of direct access to patients. I think there are incredible opportunities for ambitious, politically savvy academic pathologists to reach across disciplines and claim some of this cutting edge turf for the field. If only I were better looking... 😉

 

[superkeith]

i dont know what box this radiologist is talking about, and radiologists are not the specialists that come to my mind when it comes to thinking outside the proverbial box.

radiologists can provide the physics background, but i do not believe the average-trained radiologist possesses greater diagnostic acumen than the average-trained ap/cp pathologist in so much as i would not ask a radiologist to describe the molecular mechanism of a particular disease process and correlate this with its clinical presentation.

 

[CameronFrye]

i dont know what box this radiologist is talking about, and radiologists are not the specialists that come to my mind when it comes to thinking outside the proverbial box.

radiologists can provide the physics background, but i do not believe the average-trained radiologist possesses greater diagnostic acumen than the average-trained ap/cp pathologist in so much as i would not ask a radiologist to describe the molecular mechanism of a particular disease process and correlate this with its clinical presentation.

I'm not exactly sure what you're talking about. It wasn't too long ago that X-rays were the only tool available to radiologists. They have gone on to develop CT, MRI, PET, interventional procedures, etc. Now they are working on molecular imaging. For the most part, pathologists are still using the same main tool (H&E) that they've been using for decades.

 

[superkeith]

I'm not exactly sure what you're talking about. It wasn't too long ago that X-rays were the only tool available to radiologists. They have gone on to develop CT, MRI, PET, interventional procedures, etc. Now they are working on molecular imaging. For the most part, pathologists are still using the same main tool (H&E) that they've been using for decades.

uh...roentgen, hounsfield, phelps and ter-pogossian were not radiologists. they are physicists/engineers that developed technology with medical applications.

is the issue radiologists vs. pathologists in regards to having the capacity to, diagnostically, "think outside the box" in approaching a given clinical scenario. if that is the case, then i do not see a particular advantage toward radiologists.

is this an issue of radiologists better being able to think outside the box because non-radiologists have developed "better" tools? if that is the case, i am inclined to believe that routine laboratory results, immunohistochemistry, cytogenetics, and molecular studies, tools that have traditionally been managed or employed by pathologists, have also positively impacted patient care.

edit: i believe plain films are still being ordered.

 

[PathJet]

we both do cool stuff radiologists and pathologists

we just need to continue doing more cool stuff and maybe just maybe we path and rads can figure out how to do cool stuff together---hand in hand...😍

cant we all just get along???

 

[superkeith]

we both do cool stuff radiologists and pathologists

we just need to continue doing more cool stuff and maybe just maybe we path and rads can figure out how to do cool stuff together---hand in hand...😍

cant we all just get along???

radiologists provide essential diagnostic information. it is absolutely inconceivable to imagine a medical practice without a radiology service. my issue is with one particular radiologist suggesting that "thinking outside the box" is more the province of radiology than pathology.

the statement perturbs me, but less so now. clearly, i disagree, but this does not mean that others cannot believe it, nor that i am necessarily even correct. i think a cheeto went down wrong when i first read it.

 

[yaah]

I always get a kick out of the whole "outside of the box" phenomenon. What I see happening these days are people are doing so much outside of the box thinking that they are neglecting what's in front of them and more vital to current practice (i.e. "inside the box.").

What is going to be important in the future is not necessarily increasing technologies or advances, but how they are applied to actual practice. That's the box. What do patients and clincians want? To me, doing extensive molecular workup on things that provides you a ton of information, all of it less helpful than an H&E, is worthless if it doesn't give you anything. Developing more and more new, technologically advanced and specialized assays doesn't help much if the results aren't that clinically useful (even if they are "neat").

What I see happening now is risk stratification - diseases and entities are becoming categorized based on different factors they do or don't have, all in a hope to guide treatment, figure out prognosis, disease aggressiveness, etc. But none of it is 100% - it's all stratification based on a number of factors, which are only important in summation. Individual tests and results molecularly are with some exceptions (mostly in hemepath currently) less important than the overall picture. Now, perhaps people will find similar things in carcinomas that are relevant, but not yet!

As an example, there is a lab in CA which runs serology on patients with suspected IBD, aimed at telling you whether the patient (by expression of various antibodies) is at increased risk for crohn's or UC, or which one they are more likely to have. I won't pass judgement on it, but there is a lot of literature out there (both biased and not so biased). These are screening tests, again which are aimed at stratifying risk and whether there is indication for FURTHER testing. Lots of these tests out there aren't taking away from anything that is currently offered, they are augmenting it or causing even MORE of the current testing.

 

[Neddy]

What I see happening now is risk stratification - diseases and entities are becoming categorized based on different factors they do or don't have, all in a hope to guide treatment, figure out prognosis, disease aggressiveness, etc. But none of it is 100% - it's all stratification based on a number of factors, which are only important in summation. Individual tests and results molecularly are with some exceptions (mostly in hemepath currently) less important than the overall picture. Now, perhaps people will find similar things in carcinomas that are relevant, but not yet!

Agreed that prognosis & risk stratification can only go so far before the clinical marginal utility falls to nill. The real advances have been and will be in predictive testing - predicting response to specific targeted therapies (e.g. not 'to treat or not to treat', but 'which of ten different chemo options will work the best') I think pathologists need to be aggressive in participating in therapy-oriented research at an early stage, because these things can have a limited window of opportunity. Tissue-based diagnostics might be the cheapest, most reliable way to determine if a given tumor will respond to drug Y, but if the pivotal clinical trials for that drug stratify patients on the basis of serology, or some new bazillion dollar molecular imaging technique instead, that will become the technology that clinicians will suddenly "want". The pathologist who comes along late with a simpler and better solution will be left begging for tissue samples to do his retrospective study.