Future Path?

This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.

cartwright77

New Member
10+ Year Member
15+ Year Member
Joined
Nov 8, 2004
Messages
3
Reaction score
0
Hi all,
great forum with some flamboyant and entertaining personalities. I'm a long-time lurker/first-time poster 3rd year who has recently developed an interest in pathology. Many of my motivations seem relatively familiar: interest in doing something that directly translates science into medicine, like the "expert diagnostician" feel of path, enjoy the visual cryptography of surg path, no rabid desire for long-term longitudinal patient relations.

That said, I'm also really interested in path b/c it seems like pathology really has the potential to be at the vanguard of genomic medicine. I really think that pathology is sort of on the precipice of major change, with the convergence of a number of technologies:

1) Digital imaging - while not exactly commonly, virtual slide technology is already in place and available commercially. I'm not talking about cumbersome telepathology setups with what amounts to remote-controlled microscopes and image transmission over IP, but rather true virtual slides, with very high-resolution slide capture and storage and familiar, intuitive interface. You may be interested to check out:

http://216.204.84.52

for some sense of how this technology might work.

2) Molecular diagnostics/mol imaging - obviously these are burgeoning fields with tons of promise. This is one of the reasons I'm really interested in path: finally seeing some of the promise of the human genome project come to fruition in terms of clinical applicability.

Ultimately, I think what you'll see is that while traditional surg path will continue to exist for some time, it will become much more integrated with genomic technologies, such that you'll see things like in situ oncogene expression analysis, cell surface receptors, etc overlayed on the digital slide. Ultimately,all of this information will be simultaneously available and displayed in a coherent fashion, versus its current fragmented state. I like to imagine that it will be like that scene from "Minority Report" where Tom Cruise rapidly parses all of this information from disparate sources by moving around all sorts of data with his hands on this nifty multi-functional display. Of course, it's a bit hard to imagine pathologists dropping in from helicopters for any relevant purpose, so obviously this analogy only extends so far... 😉

Does anyone else feel the same? That AP/CP divisions are a bit archaic and artificial, and particularly with molecular diagnostics/genomic medicine becoming (relatively) commonplace, things are going to change rapidly? This is honestly one of the big draws to path for me right now, the opportunity to be involved with much of the emerging molecular technology as it makes its transition into clinical relevance.

On another note, do you guys ever wonder about Interventional Pathology? I mean, I've read and enjoyed the extensive ACM thread, so I understand that a number of the frequent posters here have little interest in treating patients per se, but would you be interested in some sort of therapeutic stuff in the context of genomic medicine, for example? I read an interesting commentary about the possibility of Interventional Path here:

http://www.labinfotech.org/LIS2005/commentary20041209.php

[Of course, any sort of proposed Interventional Pathology would probably require an internship or previous residency, which is likely to get the big 👎 here.]

Anyway, just some thoughts. I really do think path may be the way to go for me, but I am really interested in people's perceptions of how the field may change (I believe dramatically) in our practice lifetimes.

cheers,
cartwright

P.S. Any golf fans here? Pretty wild Open!
 
A lot of people think the field is changing, however radical changes such as you propose are unlikely to have a serious impact on daily practice for most over the next few decades. Molecular diagnostics is growing exponentially, but the diagnosis is still made via more traditional methods for the most part - and molecular is used to supplement diagnosis or perhaps confirm it. It will take a long while for established practices to change. The problem is that there are no absolutes. Not every Ewing's Sarcoma has the same translocation or expression. Not every AML M2 has the same karyotype. Some people believe molecular diagnostics is going to be very important, but perhaps from a prognostic standpoint - i.e. tumor is diagnosed via traditional methods and molecular studies identify which genes are active and which treatments are likely to work. It will probably lead to lots of money for the right people, particularly drug companies who will begin to individualize treatments based on genetic expression.

As for telepathology though, there is a lot of research into it but I haven't talked to many who think it is seriously going to replace traditional methods anytime soon. It is still too cumbersome and the resolution can never approach what you get from a microscope and your own eyes.
 
cartwright77 said:
P.S. Any golf fans here? Pretty wild Open!
Not to hijack the thread, but since you asked...
I was like everyone else and thought Goosen would be solid in the last round, and, once he cracked, figured Tiger would go ahead and do his thing, but that's golf...
Love to see that kind of course with the "around the green" options.
 
Yaah,
as you point out, it's highly unlikely that some sort of genomic fingerprint is going to be consistent enough to reliably replace standard histopathology for tumor dx. However, I really do believe that molecular markers are going to become integral to pathology practice, not just as the sort of adjunct that they are now. Personalized medicine, as everyone likes to call it, will definitely be important for both prognostic purposes and therapeutic interventions, once there is enough data re: expression profiles of various neoplasms.

As far as virtual slide systems not being capable of the resolution of the microsope, I would argue that this isn't strictly true with today's technology. I did a little google research on things, and at least from a resolution perspective, current state of the art slide scanners are scanning at resolutions greater than 40x objective equivalents. [i.e., optical resolution at 40x is 0.45u, and at least one system I saw scans at 0.25u/pixel]. I haven't done a ton of path at this point, but how often do you really need to go to oil to make the dx? Some interesting calculations here regarding the amount of data on a slide:

http://www.telepathologycity.com/quantifying_histological_data.htm

Yes, there are definitely some IT concerns: these scans generate massive files, on the order of 10-15 GB/slide, and obviously we're talking about serious amounts of data storage. But hey, multi-terabyte arrays of disk storage are common and inexpensive (I mean, I have a MythTV setup running Linux with 2 TB of storage at home!), and 10 Gb/s ethernet is becoming standard, so I don't really think it's all that impractical.

Personally, I think that Path is where Radiology was 10-15 years ago WRT to digital imaging. There's a lot of grumbling about it being impractical, resolution problems, storage, reliability, blah, blah, blah, but these are technical problems that are rapidly being overcome. Any place that doesn't have digital radiology/PACS (at least in the US) is largely living in the dark ages at this point.

No offense to anyone here, but there's certainly a lot of traditional sentiment in pathology and my (admittedly limited) perspective on this is that this fosters a reluctance to embrace what's inevitably coming. PathPACS offers so many advantages it's ridiculous: workflow integration, remote retrieval/collaboration, image analysis/quantification, overlay of IHC/? molecular data in the future, ad infinitum. I will be amazed if this isn't a standard in the next 10 years at any modern facility.

If you have time or interest, here is one place (among many) that offers a fully automated slide scanner into the pathology workflow, so that along with automated tissue processor/staining/coverslipping, virtual slide acquisition is just the final step in the process.

http://www.aperio.com/productsservices/prod-t3.asp

[Honestly, I'm not affiliated with any of these companies! 😉 I just think digital technology is pretty damn cool. I really have tried to make myself like rads, but having a basic science background, I think path is the way to go.]

cheers,
cartwright

US Open: amazing the score that Goosen posted! I agree, Pinehurst No. 2 is a great course and a real test of skills around the green. Fascinating to watch the pros repeatedly putt it off the green from one side to the other. Can't imagine what I'd shoot there!
 
There is a lot of tradition in path, you're right. But it's also true that almost anyone who has spent significant time using a microscope to make a histologic diagnosis, and then tries the same thing on a computer screen will prefer the microscope and probably insist that they see more and are more likely to make an accurate diagnosis. I tend to agree - computer images are harder to work with and while the resolution is good, it just isn't the same as binocular vision and proper optics. Now, if the resolution is so good, perhaps they can just create something like a computer with eye pieces and binocular vision...:laugh:
 
I won't contest the fact the user interface on the various virtual slide systems that I've seen is probably not as intuitive as it could be. Again, though, I think the rads analogy holds: 15 years ago, radiologists likely insisted that a CXR on a lightbox revealed far more subtle detail and was much easier to work with than anything that could be reproduced on a computer monitor (at the time). Once UI improves and pathologists realize the manifold advantages to virtual slide, I don't think there will be any going back.

Who knows how long this will all take, but I won't even start practicing for 6 years (4th year+residency+fellowship), so I'll definitely be looking for places that embrace the new stuff when I'm finishing.

Perhaps LADoc can finally realize his dream of signing out cases at Starbucks alongside his rads compadres...
 
cartwright77 said:
Yaah,
as you point out, it's highly unlikely that some sort of genomic fingerprint is going to be consistent enough to reliably replace standard histopathology for tumor dx. However, I really do believe that molecular markers are going to become integral to pathology practice, not just as the sort of adjunct that they are now. Personalized medicine, as everyone likes to call it, will definitely be important for both prognostic purposes and therapeutic interventions, once there is enough data re: expression profiles of various neoplasms.

As far as virtual slide systems not being capable of the resolution of the microsope, I would argue that this isn't strictly true with today's technology. I did a little google research on things, and at least from a resolution perspective, current state of the art slide scanners are scanning at resolutions greater than 40x objective equivalents. [i.e., optical resolution at 40x is 0.45u, and at least one system I saw scans at 0.25u/pixel]. I haven't done a ton of path at this point, but how often do you really need to go to oil to make the dx? Some interesting calculations here regarding the amount of data on a slide:

http://www.telepathologycity.com/quantifying_histological_data.htm

Yes, there are definitely some IT concerns: these scans generate massive files, on the order of 10-15 GB/slide, and obviously we're talking about serious amounts of data storage. But hey, multi-terabyte arrays of disk storage are common and inexpensive (I mean, I have a MythTV setup running Linux with 2 TB of storage at home!), and 10 Gb/s ethernet is becoming standard, so I don't really think it's all that impractical.

Personally, I think that Path is where Radiology was 10-15 years ago WRT to digital imaging. There's a lot of grumbling about it being impractical, resolution problems, storage, reliability, blah, blah, blah, but these are technical problems that are rapidly being overcome. Any place that doesn't have digital radiology/PACS (at least in the US) is largely living in the dark ages at this point.

No offense to anyone here, but there's certainly a lot of traditional sentiment in pathology and my (admittedly limited) perspective on this is that this fosters a reluctance to embrace what's inevitably coming. PathPACS offers so many advantages it's ridiculous: workflow integration, remote retrieval/collaboration, image analysis/quantification, overlay of IHC/? molecular data in the future, ad infinitum. I will be amazed if this isn't a standard in the next 10 years at any modern facility.

If you have time or interest, here is one place (among many) that offers a fully automated slide scanner into the pathology workflow, so that along with automated tissue processor/staining/coverslipping, virtual slide acquisition is just the final step in the process.

http://www.aperio.com/productsservices/prod-t3.asp

[Honestly, I'm not affiliated with any of these companies! 😉 I just think digital technology is pretty damn cool. I really have tried to make myself like rads, but having a basic science background, I think path is the way to go.]

cheers,
cartwright

US Open: amazing the score that Goosen posted! I agree, Pinehurst No. 2 is a great course and a real test of skills around the green. Fascinating to watch the pros repeatedly putt it off the green from one side to the other. Can't imagine what I'd shoot there!

Very cool stuff! That demonstration video was pretty impressive. Maybe one day telopathology in some form or another will be standard.
 
Top