Does pathology morphology matter anymore?

[caffeinegirl]

Greetings from the Pathology Forum! As pathologists, we have seen an increased trend towards oncologists just focusing on molecular test results rather than the standard pathology report. See below. What are you thoughts? Does morphology (ie. anatomic pathology diagnosis) matter anymore?

Digital pathology is just pathology, but does it matter anymore? | Tissuepathology.com

Since the 1990s, many innovators to Fortune 500 and Fortune 50 companies have worked to make digital pathology a reality. While not mainstream and perhaps less than 10% of US based pathology groups have any serious implementations of digital pathology, the technology has caught up to our retinas...

tissuepathology.com

---

Members don't see this ad.

 

[ShuperNewbie]

Yeah it does. Small round cells on a lung biopsy are still taken very seriously. We also use certain features to prognosticate and counsel patients (e.g. would be more inclined to offer adjuvant chemo in a premenopausal lady if her breast cancer pathology has metaplastic features).

I don't think anyone thinks path is dying anytime soon.

 

[Duke1K]

On the heme side, marrow morphology absolutely matters too. Rarely can you make a diagnosis based on molecular alone, but most often it assists with (not replaces) a morphological diagnosis and treatment options.

 

[whoknows2012]

On the heme side, marrow morphology absolutely matters too. Rarely can you make a diagnosis based on molecular alone, but most often it assists with (not replaces) a morphological diagnosis and treatment options.

Most acute leukemias are purely flow molecular or cg based diagnoses. Morphology is an afterthought mostly with one caveat being monocytic aml where morphology can be a companion if needed to molecular flow and cg based diagnostics

 

[gutonc]

Of course histology matters. But the molecular information is also important since it can guide treatment and prognosis discussions in ways that just saying "lung cancer" can't.

Just to give the oncology side of the examples the author used:

Lung mass biopsy (assuming metastatic disease) - "small cell or non-small cell" is always the first question. If it's small cell, I'm done and will make my treatment decisions based on that alone. I will probably get NGS testing later, but it's not critical to starting treatment. If Non-small cell, is it squamous or not? If it's squamous, I'm going to wait on the PD-L1 to determine chemo/chemo-IO/IO. If it's non-squamous, I'm waiting for the NGS too, since EGFR and ALK mutations (among others) do better with targeted therapy than chemo.

Liver mass - In general, we don't typically biopsy slam dunk HCCs for pathology since we can usually diagnose them using clinical and radiographic characteristics. So in that case, yes, if I were getting a biopsy, it would be NGS. On the flip side, I can count on one hand the number of times in the last 12 years I've had a pathologist say "definitely a cholangiocarcinoma" rather than "poorly differentiated carcinoma with pancreaticobiliary features, clinical correlation required". So in that case, I get the NGS because there are a number of mutations and rearrangements that help nail down the histology and also help with treatment planning.

ETA: Sometimes I'm getting a biopsy of a cancer I already know about because I'm concerned about mutation induced drug resistance and the presence of secondary mutations (we see this not uncommonly in met breast (ESR1 and PI3k/PTEN/AKT axis mutations), EGFR mutated lung (T790M and secondary MET mutations) and GIST (KIT and PDGFRa mutations). In those cases, I really don't care about the histology, because you already told me what it was, I really do just need the NGS panel.

 

[Duke1K]

Most acute leukemias are purely flow molecular or cg based diagnoses. Morphology is an afterthought mostly with one caveat being monocytic aml where morphology can be a companion if needed to molecular flow and cg based diagnostics

Fair - in my mind, IHC/flow is lumped in with morphology (since IHC is essentially a very specific histological stain, and flow overlaps significantly with IHC), while molecular testing refers to NGS/cyto. Without IHC/flow, all hematolymphoid disorders would be challenging to diagnose in this current era.

 

[caffeinegirl]

I don't think anyone thinks path is dying anytime soon.

Agree - but path is changing rapidly. There are research algorithms in place that can detect morphological phenotypes that correlate with molecular markers (and that can't be quantified with the human eye). It's just a matter of time before it changes how we diagnose and stratify malignant neoplasms.
Molecular diagnostics completely changed our field - combining morphology and molecular phenotypes seems to be the next step.

 

[gutonc]

Agree - but path is changing rapidly. There are research algorithms in place that can detect morphological phenotypes that correlate with molecular markers (and that can't be quantified with the human eye). It's just a matter of time before it changes how we diagnose and stratify malignant neoplasms.
Molecular diagnostics completely changed our field - combining morphology and molecular phenotypes seems to be the next step.

It absolutely did. And at a speed that your mentors aren't likely to have picked up on and passed on to you, or other trainees.

I am fortunate to now work in a setting where I don't have to explain the need for molecular/genomic studies to my pathologists, but at my last job it was like pulling teeth to get them to even consider the need for it.

But I don't get why you wouldn't embrace molecular diagnostics and make it a value add. I have gotten some incredibly detailed molecular results with the findings, the potential drugs that might work, references to available literature and links to open clinical trials. That's where you need to put your energy. You're not going to beat molecular testing with an H&E, but you can make it better.

 

[caffeinegirl]

It absolutely did. And at a speed that your mentors aren't likely to have picked up on and passed on to you, or other trainees.

I am fortunate to now work in a setting where I don't have to explain the need for molecular/genomic studies to my pathologists, but at my last job it was like pulling teeth to get them to even consider the need for it.

But I don't get why you wouldn't embrace molecular diagnostics and make it a value add. I have gotten some incredibly detailed molecular results with the findings, the potential drugs that might work, references to available literature and links to open clinical trials. That's where you need to put your energy. You're not going to beat molecular testing with an H&E, but you can make it better.

Ummm...I do embrace molecular diagnostics.. I included a link to an article by another pathologist. Don't knock my mentors or myself - I think you have misunderstood my post.

 

[gutonc]

Ummm...I do embrace molecular diagnostics.. I included a link to an article by another pathologist. Don't knock my mentors or myself - I think you have misunderstood my post.

I may have misunderstood you, in which case I apologize. And I was using the royal you, not you specifically.

The pathologist in the blog post you linked seems to think that we're idiots for wanting molecular data when we should be happy just getting the histology and maybe a couple of immunostains to prove the point.

I for one want all the information, histo, cyto, immuno, molecular, before making any treatment decisions. I think the biggest problem we run into now is the delays in getting all of it ordered, resulted and available in a timely manner.

I also see a lot of patients referred in from "outside hospital" where the referring PCP or surgeon will say something like "the radiologist says it looks like lung cancer, so do you really need a biopsy or can you just start chemo?". And patients who expect to be starting treatment on their first visit, literally one day after their colonoscopy, before the path tech has even cut the paraffin slides.

 

[caffeinegirl]

No worries @gutpath - I was just looking for input from the heme/onc community to get a different perspective. As pathologists, you have to embrace the multidisciplinary nature of the field - it's not just anatomy/histology now, but also prognostic and treatment markers. But I do think the author of the article has a point - after the initial diagnosis of "malignant", these molecular/treatment markers matter more than a nuanced diagnosis in the majority of cases. Our tumor boards are also leaning more towards discussion of the molecular/prognostic markers in most cases rather than the standard path report.
It's nice to get your perspective, and good to know that the combination of morphology, molecular and immunophenotype are still helpful. I do think 10-15 years down the road, AI-morphological phenotypes will also be included for treatment and prognosis.

 

[gutonc]

No worries @gutpath - I was just looking for input from the heme/onc community to get a different perspective. As pathologists, you have to embrace the multidisciplinary nature of the field - it's not just anatomy/histology now, but also prognostic and treatment markers. But I do think the author of the article has a point - after the initial diagnosis of "malignant", these molecular/treatment markers matter more than a nuanced diagnosis in the majority of cases. Our tumor boards are also leaning more towards discussion of the molecular/prognostic markers in most cases rather than the standard path report.
It's nice to get your perspective, and good to know that the combination of morphology, molecular and immunophenotype are still helpful. I do think 10-15 years down the road, AI-morphological phenotypes will also be included for treatment and prognosis.

The bolded is true, not because we don't respect your "mad skope skillz" but because the microscope can't tell me if there's an EGFR or KRAS mutation, or what the PD-L1 CPS/TPS is, and those factors directly impact treatment and prognosis.