Upcoming Events
Log In
Pricing
Free Trial

Atypical Lung Cancers, Dr. Lacey McIntosh (10-15-26)

HIDE
PrevNext

0:02

Hello, and welcome to Noon Conference hosted by Modality.

0:06

Noon Conference connects the global radiology community through free live

0:09

educational webinars that are accessible for all, and is an opportunity to learn

0:13

alongside top radiologists from around the world.

0:17

Today, we are honored to welcome Dr.

0:19

Lacey McIntosh for a lecture entitled Atypical Lung Cancers. Dr.

0:23

McIntosh is an associate professor of radiology at UMass Medical School and

0:28

specializes in cancer and molecular imaging.

0:31

She currently serves as the Assistant Program Director, Diagnostic Radiology

0:36

Residency, and the Division Chief, Oncologic and Molecular Imaging at UMass

0:41

Memorial Medical Center. She loves teaching and working with residents.

0:45

At the end of the lecture, please join Dr.

0:47

McIntosh in a Q&A session where she will address questions you may have on today's

0:51

topic. Please remember to use the Q&A feature to submit your questions so we can

0:55

get to as many as we can before our time's up.

0:57

With that, we are ready to begin today's lecture. Dr.

1:00

McIntosh, please take it from here.

1:07

Great. Thank you so much for the warm introduction. I'm so happy to be here.

1:12

Thank you for having me. So yes, today we're going to talk about atypical lung

1:17

cancer.

1:20

I have a few disclosures for clinical trial reads and curriculum creation, but

1:24

nothing relevant to this talk. So today we're going to review some historical lung

1:29

cancer facts and trends, as well as explore typical and atypical features of lung

1:35

cancer. We're going to review specifically radiologic findings of typical and

1:40

atypical lung cancers.

1:44

So when approaching atypical lung cancer, we first must define what typical lung

1:49

cancer is, but how we define typical isn't exactly hard and fast, right?

1:54

It's a fluid and relative concept, and it considers different elements.

2:01

One of these elements is time. What's typical today was probably not typical in the

2:06

past and may not be typical in the future.

2:08

What shapes our perception of typical is contextual, it's changing, sometimes

2:13

cyclical, so we must sort of reconcile the past and the future, understand the

2:18

present to reexamine what we know about lung cancer in light of evolving evidence

2:22

to redefine for us today what is typical and thus atypical.

2:28

Subject to this constant change itself is indeed lung cancer.

2:32

Just a little more than a century ago, lung cancer was extremely rare and certainly

2:37

fell into the category of atypical neoplastic diseases.

2:41

We've seen changes over time. Just diving back a little bit into the history, while

2:45

the late 1800s showed some of the first documented evidence of lung cancer, it

2:49

certainly wasn't the beginning. As early as 1500, miners in Central and Eastern

2:54

Europe developed Bergbrand, Scheinberg Mountain sickness, later identified as lung

2:59

cancer. Certain regions were identified as death pits, where virtually all workers

3:04

got sick. People started to take the time to examine causation, later attributing

3:08

the sickness to arsenic, radium, and radon gas exposures.

3:12

In the US, lung cancer reports were very rare prior to the early 1900s, and some of

3:17

the first reports were linked to uranium and mining in Colorado, which were pursued

3:21

for manufacturing atomic bombs and building a nuclear arsenal.

3:28

By the 1940s and '50s, cigarette smoking had become extremely popular, endorsed by

3:32

many, including doctors and celebrities, and lung cancer rates were on the rise.

3:36

Tobacco companies were effective in their propaganda, calling the links to lung

3:40

cancer controversial and demanding more research, despite case-controlled studies

3:44

and internal research supporting the link.

3:48

Fast-forward to today, lung cancer is the most commonly diagnosed cancer and

3:52

leading cause of cancer deaths worldwide.

3:55

We do, however, note a more recent slight decline in rates that further emphasizes

3:58

the fluidity in our perception of what is typical for lung cancer.

4:04

So in this presentation today, we're going to look at what makes lung cancer

4:07

typical and thus atypical from multiple different perspectives, including the types

4:12

of patients with lung cancer, the anatomic location and histology,

4:18

radiographic appearance, metabolic and molecular behavior,

4:23

to things like detection, management, and finally, prognosis.

4:29

We'll examine the fluid and dynamic aspects of each of these.

4:33

So let's start with patients we see with suspected or proven lung cancer in our

4:37

clinics. This is who we would consider a typical patient with lung cancer, an older

4:43

cigarette-smoking male. However, we are also encountering increasing numbers of

4:48

atypical patients with lung cancer.

4:53

Included in these groups are those with secondary exposures.

4:57

An interesting group of patients we see with lung cancer are non or low smokers

5:01

with asthma who develop carcinoid tumors, itself an atypical lung cancer with a

5:06

potential paradoxical protective effect from smoking.

5:10

We also see patients with genetic predispositions, those with a family history of

5:14

lung cancer, and females who have unique sex-specific risks.

5:19

In particular, a group of East Asian female never smokers with lung cancer has been

5:23

receiving increasing attention.

5:26

We can see in this graph from a recent study in Taiwan showing that despite stable,

5:31

low smoking rates, we have a notable increase in lung cancer incidents over

5:37

time, which is mainly driven by adenocarcinoma.

5:42

Finally, as patients live longer with better outcomes for their lung cancer, it's

5:47

becoming less atypical to diagnose second, third, or more lung cancers in the same

5:51

patient.

5:55

Let's now discuss the anatomic location and histology of lung cancer.

6:01

We've certainly noticed a shift in the typical anatomic location of lung cancers,

6:06

primarily triggered by changes in cigarette composition and smaller particulate

6:10

pollution. Early filterless cigarettes deposited particles centrally in the airways

6:16

and typically caused more central lung cancers.

6:19

The introduction of filters in cigarettes changed this pattern.

6:22

Smokers needed to inhale more deeply and forcefully, driving deposition deeper into

6:27

the lung, with filters blocking the larger particles, but allowing smaller

6:31

particles to travel further into the lung.

6:34

This resulted in more peripherally located lung cancers.

6:37

So we can see now that what was previously atypical has become typical with this

6:43

peripheral distribution.

6:48

As for histology, the three main types of typical histologies

6:55

compromising the majority of lung cancers that we see are usually adenocarcinomas,

7:00

squamous cell carcinoma, and small cell carcinoma.

7:03

These three together make up approximately 80% of lung cancers that we see in our

7:08

clinics. However, there are many less typical histologies such as carcinoid,

7:14

large cell, sarcomatoid carcinoma, which have emerged, shifting this distribution

7:19

of histologies. Both the increasing numbers and resulting shifts should raise our

7:23

awareness for these atypical lung cancer histologies.

7:30

Molecular profiling of lung cancers has become increasingly important.

7:34

Just two decades ago, this was neither typical nor atypical, but simply

7:39

nonexistent. However, today, gene sequencing is of paramount importance, not only

7:44

for the diagnosis, but arguably more important for the selection of targeted

7:49

treatments, and this has become a typical mainstay in the workup of lung cancer

7:54

today.

7:57

With that, we come to the central part of this presentation regarding the

8:02

radiographic appearances of lung cancer.

8:07

Typical appearances of lung cancer usually consist of a nodule that is solid,

8:13

sub-solid, or part solid. Spiculation and pleural retraction are common features

8:19

of malignancy. You can see here this large tumor, this mass has a lobulated

8:26

contour with maybe some areas of microlobulation.

8:29

The next panel shows a mixed solid and ground-glass component nodule, and then the

8:35

third shows a mostly solid nodule with definite spiculation and some pleural tags

8:41

at the inferior aspect. And then the last two show pleural retraction and visceral

8:48

indentation with spiculation in the last panel.

8:55

But not all cancer follows these rules, such as this atypically benign-featured

9:00

nodule, which demonstrates a rectangular shape with fairly smooth and well-defined

9:05

margins However, on follow-up, this nodule later declared itself malignant

9:11

and was found to be a small cell carcinoma.

9:14

You can see the transformation from somewhat benign features to

9:20

microlobulated and lobulated contours with spiculations and pleural indentation at

9:26

the right lateral aspect.

9:31

We also see here a typical irregular cystic lesion, which over time showed

9:36

increasing complexity, which was also found to be malignant.

9:41

This area of cystic lung cancers is emerging and

9:46

increasingly being incorporated into our lung cancer screening nodule

9:51

management guidelines, which is very important because these are sort of newly

9:57

recognized as being malignant potential.

10:03

We see these atypical complex cystic lung cancers which show varying degrees of

10:07

cystic and solid components. The lesion on the left I would describe as

10:12

predominantly cystic with some areas of wall thickening, some of which is smooth,

10:17

where the yellow arrow is, and some of which is more nodular and irregular, where

10:21

the red arrow is. The lesion on the right is, I think, more complex, but

10:26

predominantly solid with prominent cystic components, but this also demonstrates

10:31

features of spiculation and irregular margins as well.

10:37

Cavitation in a nodule can be suggestive of infection, but can also be seen with

10:43

malignancy, as was the case for this small squamous cell carcinoma.

10:52

This left lower lobe consolidation has a typical appearance of pneumonia.

10:56

We see consolidation with air bronchograms on the first image, and then we can

11:02

see an increase in these findings on the second image.

11:05

This did not respond to antibiotics and actually turned out to be a mucinous

11:10

adenocarcinoma. Non-mass-like cancers mimicking infection are atypical in

11:15

appearance.

11:19

Lung cancers that occur in the setting of background architectural distortion can

11:23

also appear atypical, as in this case of an intensely FDG avid right upper lobe

11:29

lung cancer. This demonstrates a finger-like projection at the right lateral

11:33

aspect, which was due to growing around areas of bullous emphysema.

11:42

And perhaps the most atypical presentation of lung cancer is when it arises within

11:46

hilar or mediastinal lymph nodes and without an actual primary lung lesion, as we

11:51

see here on this PET-CT.

11:57

Beyond static morphology, we must also examine dynamic morphology or changes in

12:02

morphology over time.

12:05

Lung cancer typically grows in two patterns, either aggressive and growing quickly

12:10

or more indolent, slow-growing spectrum lesions that develop over years, with

12:14

changes in density sometimes being more notable than changes in size.

12:19

Atypical growth patterns may include long periods of stability followed by rapid

12:24

growth,

12:25

hyper-aggressive growth and spread, which sometimes can be confused for infection

12:31

because it's moving so quickly, or simply unexpected or unusual changes in

12:36

behavior.

12:39

Let's look at a few examples. So this type of unexpected dynamic morphology is

12:44

depicted here, where we have an atypically behaving malignant nodule, which showed

12:50

a relatively long period of stability with sudden rapid growth.

12:53

You can see between 2021 and 2023, it's really minimal change in this small,

12:59

mostly solid nodule. But between the same two-year time interval between 2023 and

13:05

2025, we have rapid increase in size of this nodule, which is

13:11

definitely more doubled in size, maybe more like five times or six times its

13:15

original size.

13:17

So this is sort of atypical dynamic morphology for malignancy.

13:25

Here we have a case of a lung cancer patient who experienced a delay in care during

13:30

workup and required a new baseline PET after two months.

13:36

And unfortunately, this showed atypically aggressive, widespread metastases.

13:42

So this is what falls in that category of sort of hyper-aggressive growth.

13:49

And finally, we have a case of typical and then atypical appearance of what was a

13:54

biopsy-proven hamartoma,

13:57

which had been stable over many years leading up to 2019, but then in 2025 started

14:02

showing interval growth, more lobulation at the margins, and then on soft tissue

14:07

windows, you can appreciate this aggressive chest wall invasion, which reflects a

14:12

malignant degeneration or transformation of this nodule.

14:18

Multiplicity of nodules or masses is typically suggestive of infection or

14:22

metastases and isn't usually indicative of a primary lung cancer.

14:27

But sometimes we do see multifocal cancers affecting the lung of similar or

14:32

differing histologies.

14:36

These images are all from a single CT exam in one patient.

14:40

While not exactly typical, it is becoming less rare to see multifocal synchronous

14:45

primary lung cancers, particularly in smokers, due to field effect, as we see in

14:50

this patient with multifocal synchronous primary adenocarcinoma spectrum-type

14:54

lesions. We really sort of run the gamut here with lesions

15:00

that look to be mostly

15:03

ground-glass density versus the first panel in the top right shows

15:08

predominantly ground-glass, but with a solid component at the right lateral aspect.

15:13

And then the bottom row is showing sort of increasing punctate solid components,

15:17

and then over to the last two panels, we see increasing solid component, and the

15:23

very last panel is predominantly solid with spiculation.

15:27

So

15:29

I wouldn't call this the typical presentation, but I think it's becoming less rare

15:34

to see, as patients live longer, multiple lung cancers diagnosed in the same

15:39

patients.

15:44

Here's another atypical presentation of multifocal synchronous primary lung

15:48

cancers, but in this case, we have lesions that have differing cystic and solid

15:54

features, which are suggestive of varying histologies.

16:02

Another case of an atypical lung cancer is this adenocarcinoma that was

16:06

incidentally detected on a staging workup for metastatic bladder cancer.

16:11

So you can see this is a mixed solid and ground-glass nodule, which is in the

16:17

background of multiple more solid and well-defined metastases throughout

16:23

the lungs that you can appreciate on the MIP here on the right.

16:31

Here's another example of these atypical bilateral oddly symmetric upper

16:37

lobe lung adenocarcinomas. Sort of interesting that we see them both on the same

16:42

axial slice. These were incidentally detected on a staging PET for a left

16:47

ureteral malignancy with retroperitoneal adenopathy.

16:53

You can see the findings here.

16:58

So important take-home points regarding the radiographic appearance for static

17:03

morphology. There really is no one-size-fits-all model to be able to identify all

17:08

lung cancers. Benign features don't necessarily indicate benign histology.

17:14

We must remember that cystic lesions can also be malignant.

17:18

Lung cancers can have unexpected or bizarre appearances, especially when occurring

17:23

in the setting of altered background architecture.

17:27

We must also consider dynamic morphology.

17:30

Lung cancers do not always follow the expected growth and behavior trends.

17:35

Finally, multiplicity. The presence of multiple nodules doesn't always suggest

17:39

infection or metastasis. It's becoming more common to see multiple cancers

17:44

affecting the lung of same or different types.

17:51

All right. Now let's look at metabolic and molecular behavior of lung cancers.

17:58

This is a typical metabolic behavior of a left lower lobe cavitary lung cancer on

18:03

FDG PET-CT. We see a large mass with intense tracer uptake corresponding with solid

18:09

tissue. This patient happens to have a cavitary component to their cancer.

18:17

As camera technology improves, we tend to apply FDG analysis to smaller lung

18:21

cancers. However, uptake in these smaller cancers might be very low or not

18:27

perceptible at all, in stark contrast to the typical uptake shown in the previous

18:31

slide. It's important to account for small size and to adjust avidity expectations

18:37

during PET interpretation of small nodules, particularly when they demonstrate

18:41

suspicious CT morphology, as shown in this case. This is a very small nodule.

18:45

I think it measured about six millimeters.

18:48

And you can see on the FDG PET that there's appreciable but very low-level FDG

18:53

uptake.

18:55

We sort of have to use the small size to adjust our avidity expectations, because

19:01

smaller things are just going to be less FDG avid.

19:04

They have less cells to be able to accumulate the tracer

19:08

to be able to create the signal on PET.

19:11

So if you have something that looks very suspicious, this is, in particular, for

19:16

small findings, small pulmonary nodules.

19:18

If you have something that looks suspicious by CT, don't let a negative or near

19:24

negative PET talk you out of your degree of suspicion.

19:27

You have to sort of know what to expect, and a low level of FDG uptake for a nodule

19:32

this small is sort of expected.

19:38

This atypical behavior is further illustrated by the lack of FDG avidity in the

19:42

solid component of this part solid lung cancer, which by CT remains highly

19:47

suspicious. We have

19:50

spiculated margins. This is a solid component here with some subsolid surrounding

19:56

area. This solid component is big enough that I would expect to see at least a

20:01

little bit of FDG uptake on PET, but alas, we don't.

20:05

So although based on the size of the solid portion, we would expect FDG avidity, we

20:11

don't see any on the 6 AM.

20:17

Atypical behavior on PET-CT is not limited to small lesions.

20:22

We see here this large right upper lobe lung cancer.

20:25

FDG is atypically very low. You can see it's not negative.

20:31

There is some minimal FDG uptake, which is similar to perhaps slightly less than

20:36

blood pool that's seen in the adjacent mediastinal structures.

20:40

This is a mucinous adenocarcinoma, which is an entity known to potentially cause a

20:44

false low or negative

20:47

FDG PET. The reason being that the tumor gets sort of diffused by mucin

20:52

and there are relatively less tumor cells and a lot of mucin, and so you can tend

20:58

to get very low signal in these.

21:04

This is also seen here in this companion case, the CT images of which I showed you

21:08

before. Again, only minimal FDG uptake is seen in this large left lower lobe

21:14

adenocarcinoma. Again, illustrating that atypical metabolic behavior is not limited

21:19

to small cancers, but can also be seen with large cancers.

21:27

Atypical behavior on PET-CT can sometimes prompt alternative diagnoses.

21:31

As we see here, we have a fairly decent-sized right upper lobe solid mass,

21:38

which really doesn't demonstrate very much FDG uptake.

21:41

It's very similar to background tissues.

21:44

This turned out to be a carcinoid rather than one of the typical histologies that

21:48

we've already discussed. For typical or well-differentiated carcinoids, the lack of

21:53

FDG uptake is indeed actually considered typical, whereas it would be considered

21:59

atypical for other lung cancers of a similar size.

22:03

Well-differentiated neuroendocrine tumors tend to be lowly

22:08

metabolically active. They have low Ki-67s, which is a proliferation index.

22:14

They're not using a lot of glycolytic metabolism, which is why they don't have a

22:19

lot of FDG uptake.

22:23

This is further illustrated here with this small, well-differentiated left lower

22:27

lobe carcinoid, whereas carcinoid is only faintly FDG avid, as you can see in the

22:32

middle panel with a very low SUV max.

22:35

It is intensely dotatate avid on the panel on the right, which has a

22:41

higher SUV max. So this is actually a typical appearance for

22:47

a well-differentiated carcinoid tumor, being lowly FDG avid or absent FDG avidity,

22:53

and at least moderate to intense FDG uptake on dotatate.

23:01

However, to make things more confusing, as this talk is about atypical lung

23:05

cancers,

23:07

I wanted to show you this example where not all carcinoids, well-differentiated

23:11

carcinoids, will show this typical dotatate avidity.

23:15

So in this right upper lobe carcinoid, biopsy proven to be well differentiated, the

23:20

dotatate avidity is almost completely absent.

23:23

It's very hard to appreciate anything here.

23:27

So this may be explained by the tumor escaping the sensitivity on the tracer.

23:32

But sometimes it can suggest an atypical histology or malignant degeneration where

23:37

the cells are losing their somatostatin receptor expression that are used to bind

23:42

the tracer. So this case further illustrates the fluidity between typical and

23:47

atypical metabolic behaviors that are not limited to more common lung cancer types

23:52

or even FDG, but can occur in rare types of lung cancer and using different

23:57

tracers.

24:00

This fluidity is likely to become even more prevalent as we introduce novel PET

24:05

tracers, such as FAPI, pictured here.

24:09

We may learn new or unexpected patterns of behavior in lung cancers, I think

24:13

similar to what we are currently learning about with prostate cancer from PSMA.

24:22

From a practical perspective, the following should be noted.

24:25

We must acknowledge the limitations of PET for lesion characterization purposes,

24:30

especially in atypical cancers.

24:33

As atypical cancers become more common, we should be aware of

24:39

over-reliance on the expected or typical PET manifestations.

24:44

In these cases, we can't read PET-CT in isolation, but rather integrated with our

24:50

diagnostic CT, notably when typical suspicious findings are

24:56

present, and we shouldn't delay workup based on a potentially false negative PET.

25:02

Finally, we should keep in mind that atypical behavior on PET can suggest

25:06

alternative differential diagnoses.

25:12

Shifting gears, let's talk about the detection of lung cancer.

25:18

How lung cancer is detected is also certainly shifting and changing over time.

25:23

In the recent past, lung cancer was typically detected on imaging done in a

25:26

symptomatic smoker or incidentally found on imaging

25:32

performed for another reason. The rollout of lung cancer screening programs and

25:37

other lung cancer initiatives have drastically changed this landscape.

25:41

We're seeing major impacts on stage shift and outcomes, as seen again from this

25:45

study in Taiwan. You can see that currently, or at least as of 2019, there's a

25:50

complete shift in the stages where the majority are now found in early stage,

25:56

versus just not that long ago, 20 years ago in 2006, the opposite was

26:02

true, that most lung cancers were being detected in their later stages.

26:08

These atypical but emerging detection techniques are under promising investigation,

26:13

but not yet routinely incorporated into clinical practice.

26:21

Looking at global trends in detection over the last 30 years show a typical overall

26:25

increase in incidence. This is largely driven by population growth and aging, but

26:30

with interesting and somewhat atypical decreasing trends in age-adjusted rates in

26:36

higher sociodemographic index groups. So this is really interesting, right?

26:40

While the overall incidence is increasing, we're seeing certain areas where you

26:46

have age adjustments that are showing decreases in

26:50

rates of lung cancer. This is likely reflecting the impact of screening.

26:57

So screening, which once seemed ahead of its time, is now a typical cornerstone of

27:01

modern lung cancer care. Not to say that this is being utilized to its full

27:06

potential, as I'll talk about a little bit later, that this is

27:12

woefully underutilized in many communities where it's available.

27:20

So moving on to management-

27:26

The typical mainstays of treatment are dependent on stage, location, tumor,

27:32

and patient characteristics. For some patients, we may need to employ atypical

27:37

strategies, including empiric therapy and following known or highly suspected

27:43

cancers. This, I think, is happening as patients age and

27:49

live longer. Some very slow-growing lung cancers, we may

27:54

decide that patients are more likely to die with that lesion than from that lesion,

27:59

and that initiating treatment for that particular cancer may be more harmful.

28:06

The harm may outweigh the benefit.

28:09

We also are needing to employ strategies regarding volume preservation,

28:15

especially for these patients with multiple lung cancers.

28:18

We can't just keep doing lobectomies because the patients will run out of lungs,

28:22

and obviously that's not going to be a sustainable strategy.

28:26

So even when the standard of care may be a resection or a

28:31

lobectomy, we're employing different strategies like SBRT or focused

28:37

radiation, just to be able to preserve lung volume when multiple cancers need to be

28:43

addressed. Ablation is another local strategy that can be

28:49

good for volume preservation.

28:52

One scenario that's becoming less atypical is the need for management and

28:58

recognition by radiologists, but also management of treatment-induced side effects,

29:03

particularly with the immune checkpoint inhibitors and targeted therapies.

29:08

Years ago, this wasn't something that

29:11

was occurring because we weren't using these drugs.

29:13

But now both radiologists and clinicians need to be aware of these because they

29:19

have impacts on patient management and prognosis as well.

29:29

So finally, moving on to prognosis.

29:36

So prognosis is variable by stage and type, and certainly has considerably changed

29:42

over time, really heavily influenced by advances in detection and treatments.

29:48

But as recently as the 1950s, we can see that lung cancer was typically a

29:54

deadly disease, with all-stage survival rates less than 10%.

30:00

We can see compared to up to 55% today, with rates as high as 60% to

30:05

90% in early-stage cancers. So today, while the incidence continues to

30:11

increase, so does survival, coupled with decreasing morbidity and mortality rates.

30:15

So today, in contrast to the 1950s, it's atypical for a patient to die

30:21

from an early-stage lung cancer, which is really great.

30:28

Okay, we made it. We made it around the circle of the lungs here.

30:32

So

30:34

now that we've examined present-day lung cancer from many angles and discussed what

30:38

makes cancers atypical as opposed to

30:42

more typical, we have looked at these from all different perspectives.

30:47

So what are the take-home points?

30:49

The take-home message is that lung cancer can be considered atypical for a variety

30:54

of reasons as to who it affects. While older male smokers are the dominant

30:59

demographic, other groups have emerged with unexpected and sometimes unexplained

31:04

risks. So be aware of atypical patients. How it manifests.

31:10

Histology and location have changed over time, in part due to changes in smoking

31:15

and exposure, but also environmental

31:19

factors such as pollution. So don't forget about unusual

31:25

histologies and molecular profiling is now mainstream.

31:30

It can be atypical in what it looks like. Radiographic appearance can be atypical.

31:35

There really is no one-size-fits-all model, so we have to pay attention to both

31:41

static and dynamic morphologies.

31:45

Lung cancer can be atypical in how it behaves.

31:49

Metabolic and molecular behavior of lung cancers can be atypical.

31:53

Use caution in over-reliance on typical PET findings and make sure to read in the

31:57

context of findings from other imaging.

32:01

It's particularly important with chest CT.

32:05

So it can be atypical in how we find it.

32:09

Screening and early AI programs are increasing detection, and we're seeing already

32:14

a tangible impact on prognosis and outcomes.

32:18

It can be atypical in how we address it.

32:21

Management is evolving, and we sometimes must operate outside the usual algorithm

32:25

in atypical patients and scenarios.

32:30

And it can be atypical in how it looks for patients.

32:34

Typical outcomes for lung cancer have drastically improved, especially in the

32:38

recent decades, and especially as we find cancers smaller and earlier.

32:46

Now that we've looked at lung cancer today, what does the future hold?

32:50

So regarding patients and trends, incidence is still increasing,

32:56

mainly due to population growth and aging.

33:00

We're starting to see some interesting trends in certain groups

33:06

where

33:08

age-adjusted rates are falling in high sociodemographic index regions, where more

33:13

resources are dedicated to early detection and treatment.

33:17

There may be strong preventative initiatives, tobacco taxation,

33:23

control of environmental risk factors like asbestosis and radon mitigation in

33:28

homes.

33:31

And as the global smoking rates decline in developed countries, the incidence of

33:36

lung cancer in individuals who have never smoked is stable or increasing.

33:41

We'll have to see what that looks like if it's more of a redistribution or shift

33:45

versus true increase.

33:51

I think the development of lung cancer will continue to be influenced by some fluid

33:56

environmental and patient risk factors.

34:00

What will we see with increasing pollution?

34:06

As an optimist, I'd also like to ask the question about what we could see if we

34:10

make improvements in our air quality.

34:13

Will we learn anything from

34:16

during the pandemic where a lot of

34:19

factories and industry was shut down and we saw some rebuilding of the ozone layer

34:26

and

34:28

decrease in particulate pollution?

34:33

Regarding vaping-associated lung cancers, the research right now is pretty

34:37

equivocal, but there is reason for concern regarding the oncogenicity of these

34:42

products. Lag times aren't really mature yet, so I think this remains

34:48

to be seen.

34:50

What about lung cancers related to marijuana smoking?

34:53

I mention this as recreational use has become legal in many US states, so I think

34:58

we have yet to see what the impact of that will look like.

35:04

And we're starting to see evidence of microplastics being identified in every area

35:08

of the human body, so I think we need to find out the implications in the lung for

35:13

both lung cancer and chronic interstitial lung diseases.

35:23

And then in terms of detection, as I mentioned, increases in

35:29

utilization of lung cancer screening will certainly help with finding more cancers,

35:34

finding them earlier, having better prognosis for patients, but utilization still

35:40

remains really low. In

35:43

2010, of eligible patients, only about 4% were receiving lung cancer screening

35:48

rates. This is in the US. And then

35:52

as of 2024, which is the latest data we have, it's still under 20%, which means

35:57

that over 80% of eligible patients that could be getting screened for lung cancer

36:02

are not getting that.

36:06

We also need to, as we develop these global screening programs, which are pretty

36:11

far ahead in the US, but not necessarily worldwide, we need to balance appropriate

36:18

screening guidelines and inclusion into programs with what is appropriate for

36:23

each population. The inclusion in

36:28

East Asian countries may be different than it is in the US or in other

36:34

countries. So I think we need to balance developing these population-specific

36:40

guidelines, but also Keeping in mind global harmonization so that if

36:46

programs look too different from each other, we can't really compare or combine

36:50

data to really be able to learn about this on a global level.

36:55

So I think balancing appropriate guidelines with global harmonization is important.

37:03

We're also seeing these really interesting,

37:07

often AI-driven risk prediction models that are coming out.

37:12

Sybil is just one of many, but

37:16

Sybil is a deep learning model that predicts future lung cancer risk from a single

37:22

low-dose chest CT, which is pretty cool.

37:25

There are many other models that may integrate imaging, but also talk about or

37:30

incorporate family history, smoking history, genetic predisposition,

37:37

all of the epidemiologic factors that go into lung cancer as well.

37:41

So I think these are going to continue to get better and really may change the way

37:47

that we screen patients or

37:51

are able to identify cancers earlier, which is the goal.

37:58

So prevalence is increasing as patients live longer and some patients

38:03

are developing more than one cancer.

38:08

Morbidity and mortality are decreasing as we're able to treat these earlier with

38:12

newer, safer surgical techniques, as well as radiation techniques.

38:18

I think treatments are going to continue to evolve, including the expansion of

38:23

immunotherapy and potential role for diagnostics.

38:27

We're starting to see this kind of emerge in other areas.

38:30

I wouldn't

38:32

be surprised at all if this is going to be important for lung cancer as well.

38:37

So

38:38

as we get more specific,

38:42

there may be individualized treatment profiles that take into account multiple

38:47

factors that we're able to learn about with specific patients and their molecular

38:51

profile of their tumors. So it may become less algorithmic and just sort

38:57

of following different boxes and down different paths, but really more

39:01

individualized to each specific patient with each specific tumor.

39:09

And so while incidents will likely continue to increase, I think we can hope that

39:14

outcomes will also continue to improve.

39:16

This is just some projection modeling of where we could be in 2050.

39:25

But yeah, just in summary, the advances in diagnostics and therapeutics have really

39:30

rewritten what we consider typical for lung cancer today.

39:35

What we consider typical today probably won't be typical in 20 or 40 years.

39:41

It's really highly dependent on a number of factors, some of which we discussed

39:44

today, but some of which probably have yet to be realized or discovered.

39:49

So, I look forward to updating this lecture as things change in the coming years.

39:57

So I have some references listed here, but that concludes my talk.

40:02

I did want to thank Dr. Alex McGeer for assistance in refining this talk, but I'd

40:07

be happy to take any questions.

40:13

Yeah, thank you for sharing that lecture with us today, Dr. McIntosh.

40:16

At this time, we'll open the floor for any questions from our audience.

40:19

You can submit your questions through the Q&A feature.

40:26

All right. I see one question here. What is the indication of biopsy and PET CT?

40:32

What are the situations?

40:36

Let's see. So I think this is sort of practice and region-dependent,

40:42

but

40:44

I think that CT really drives a lot of this.

40:49

And so if you have a nodule that has clearly malignant features like

40:54

spiculations or

40:58

lobulated, microlobulated contours, as we talked about with all the static

41:02

morphology features. But also if you have something that has dynamic features that

41:06

are concerning, like increases in size or density over time,

41:11

those are all indications for biopsy.

41:13

PET CT is sometimes used for lesion characterization, but oftentimes, at least I

41:19

find in my practice that

41:23

we already may have a suspicion about the nodule, and the role of the PET is really

41:27

to look at if there's nodal involvement or distant metastatic disease.

41:34

Sometimes, though, we do use PET to characterize a nodule if we're not sure if

41:38

something is

41:40

malignant or benign. An example would be rounded atelectasis.

41:45

We have a pleural-based nodule or mass that we're really

41:51

not sure if it's rounded atelectasis or malignancy, then PET CT can be helpful in

41:55

sorting those two things out.

41:58

So those are just some scenarios that might drive biopsy versus surveillance or

42:03

further workup.

42:06

All right, I see another one. Can PET scan be used as a screening tool?

42:11

So right now, at least in the US, it's not really used that way.

42:15

There is a lot of excitement about whole body MRI, which would be a

42:21

screening tool. And in some places, there are programs where you

42:27

can sort of buy a PET CT for yourself.

42:31

You can do a self-pay as a screening exam.

42:36

So there are isolated cases of that, but right now it's not really being used that

42:40

way. As a person who reads primarily PET CT, I think that

42:46

you really need a high-quality CT to interpret your PET,

42:53

which depending on the specific protocols or cameras that you have available to

42:58

you, you may or may not have the ability to have a high-quality CT along with a

43:03

PET. So I think PET's really a sort of a

43:09

tool in the toolbox that helps us work out findings that we see on other imaging.

43:15

That's where I see its primary use.

43:21

Dr. McIntosh, if you're looking for the Q&A menu,

43:25

there's a dropdown there next to the

43:31

settings for your mic and stuff, I believe.

43:34

I see. Thank you. I was looking in the chat.

43:38

Let's see. Okay. I'm sorry, I didn't mean to ignore what was going on here.

43:44

Do you think it is likely Pop Leo has atypical lung cancer?

43:47

I believe he's a non-smoker. So yeah, I think the lung cancer arising in

43:54

a non-smoker is certainly atypical, but becoming more recognized.

43:58

I saw something yesterday. I think they've released information that it's an

44:02

adenocarcinoma, which is one that certainly can be associated with non-smokers.

44:08

I would say I probably, in my personal experience, haven't seen many cases of

44:11

squamous cell that have been outside of smoking population. So, okay.

44:19

Please, what is the role of MR spectroscopy in lung cancer or malignant nodules?

44:25

This is a great question. I do not know.

44:30

I don't read any MR in my current practice, but

44:35

the ability to sort of differentiate different types of tissue by MR spectroscopy

44:41

is certainly exciting and promising.

44:43

So, I'm sorry I don't have an answer for you, but I'm sure it's worth looking into.

44:48

There must be papers out there about it.

44:51

Okay. Should non-smokers be screened for atypical lung cancers, though?

44:56

That's a great question. I think right now the screening guidelines are really

45:02

heavily driven by smoking history, but as we discussed,

45:07

we see lung cancer in non-smokers, so there has to be something there specifically

45:11

or particularly in these East Asian females.

45:14

So I think when we're creating screening programs, those things are going to have

45:20

to be incorporated for certain populations.

45:23

So,

45:25

if there's a strong family history or environmental factors like pollution,

45:31

then maybe those things should be considered in designing lung cancer screening

45:35

programs and guidelines.

45:38

Let's see. Frequently, we encounter tiny nodules and pass it off as healed

45:43

granuloma. So when to suspect and when to follow up for malignancy?

45:47

Yeah, this is really tricky as we sort of deal with one nodule at a time.

45:54

One thing that can be really helpful is if it's calcified, those tend to be benign.

45:59

And so if there's calcification within a small nodule, we tend to just kind of put

46:04

that in that bucket. And they don't necessarily require further follow-up.

46:09

We use the lung cancer screening

46:13

lung-RADS as a guide for us as well, when we're formally screening for lung cancer.

46:19

And when we're not and we find

46:21

lung nodules on CT, we use the Fleischer Society guidelines.

46:25

Both of those classify

46:28

calcified nodules as being benign.

46:30

So the guideline systems that we have in place can be really helpful for sort of

46:36

managing these. So

46:39

when we think something's a granuloma, we kind of stop there.

46:43

Let's see. "From your

46:49

experience, what steps can be taken to distinguish between infection versus

46:53

malignancy in CT lung, especially if surrounded by consolidation opacity with air

46:59

bronchogram?" Yeah, this can be really tricky.

47:03

I think features like,

47:06

as we talked about today, if something is growing in a mass-like configuration,

47:11

that tends to be more concerning for cancer versus if it's just filling alveolar

47:15

spaces, that tends to be more consistent with infection.

47:20

But in particular, that left lower lobe adenocarcinoma that I showed you guys,

47:24

really it is actually just cancer that's filling the alveolar spaces.

47:30

So some clues that you can use to differentiate between these two, and I think we

47:34

miss these all the time because we just say that they look like pneumonia, is if

47:38

they don't respond to antibiotics and if they continue to get worse over time.

47:42

These are frequently diagnosed on bronchoalveolar lavage because we're not able to

47:47

identify them by imaging because they look the same.

47:49

But if they're not improving, that may direct you to recommend

47:55

lavage to look for something that's atypical, like cancer.

48:00

Let's see. "Is there

48:06

in the US something like criteria Fleischer Society guidelines to follow up the

48:10

unclear lesions?" So yeah, I touched on the Fleischer Society guidelines

48:15

before in that

48:20

the guidelines work very well if your lesions fit into a box,

48:26

right? Is this a solid nodule? Is this a calcified nodule?

48:30

Is this a cystic nodule, which isn't completely addressed yet.

48:34

But when things fall into those boxes, then there's answers of what to do with it

48:38

next. Does this need follow-up? If it does, what's the time interval that we need

48:42

it? How far should we follow it out?

48:44

But these unclear lesions are a little bit tricky, right?

48:49

So I think if there's a chance that something could be malignant, it's reasonable

48:54

to continue to follow it. Use everything that you have available to you.

48:58

Look at static morphology, dynamic morphology.

49:02

Certainly, if there's any changes over time.

49:05

What's very interesting is when we think about changes over time, we tend to think,

49:10

"Oh, okay, if a nodule is increasing in size, that's suspicious, right?" But

49:15

they're not common. There are cases of nodules that,

49:21

particularly ones that start out as sub-solid or ground-glass, and over time,

49:26

as they get more solid, the solid tissue in the

49:32

center can sort of retract, and pull the tissue in around it.

49:37

And so sometimes you have a nodule that actually shrinks over time.

49:42

But if it's increasing in density, then you sort of have to balance those two, and

49:47

that's a suspicious finding.

49:50

Let's see. "Because it's often hard to differentiate between

49:55

benign nodules, granulomas, and cancer." Yeah.

49:59

So we talked about this one a little bit.

50:01

And I think that's the whole point of these Fleischer Society guidelines, right?

50:04

Is that when we see a nodule, we often don't know.

50:08

If it's calcified, we can feel good about that.

50:10

But sometimes granulomas are non-calcified, and so we don't know, is this a

50:14

non-calcified granuloma? Is this a benign nodule that's occurring for some other

50:19

reason, or is this the earliest manifestation of lung cancer?

50:22

Because they pretty much all start as a small nodule,

50:26

or an area of increased density.

50:29

And so when we're looking at these in isolation, especially the first time that we

50:33

find them, the first CT a patient has, it can be really difficult to differentiate

50:38

them, which is why we have these systems of follow-up.

50:42

And we sometimes stratify the follow-up based on just what it looks like on

50:47

imaging. But other times we use clinical information, like with Fleischer Society

50:52

guidelines, there's different

50:55

algorithms or pathways for patients that are considered high risk versus low risk.

50:59

And smoking is

51:02

something that's considered in that.

51:05

All right. "Does a lung lesion limited by a fissure indicate it to be a more benign

51:10

lesion than a malignancy?" This is a great question.

51:14

So if you have a nodule that's located within a fissure, especially if it

51:20

has a triangular shape or concave margins to it, rather

51:26

than bulging or convex,

51:30

that is a very good appearance for an intrapulmonary lymph node, which is going to

51:35

be a benign lesion.

51:38

But you can certainly get lung cancers that arise adjacent to a fissure or the

51:44

pleural surface. And one feature that I'm always looking at when I find a

51:49

pulmonary nodule is, does this have any impact on the fissure or the pleural

51:55

surfaces nearby?

51:58

Because nodules that are benign tend to just grow in a

52:04

little mass-like configuration.

52:06

They push things that are nearby out of the way, versus primary lung

52:11

cancers are very fibrotic, and they tend to drag the tissue in around them, which

52:16

is why they become spiculated. Those spiculations are reflective of central

52:20

fibrosis and retraction. And so if you have a nodule that's on

52:26

or next to a fissure and you see any sort of fissural distortion where it's being

52:31

tented, pulled, any sort of architectural distortion there, that is a red flag for

52:36

a cancer. So that's certainly a feature that would head me in more the direction of

52:42

a malignancy.

52:46

Let's see. There's a question here about biopsy, which I won't take because I don't

52:50

do any interventions myself, so I won't be able to answer it with any expertise.

52:55

But, "When suggesting an adenocarcinoma diagnosis, do you specify AAH,

53:02

AAIS, MIA, or leave it non-specific?" I love that somebody asked this question

53:07

because this is a teaching point that I really drive home with our trainees here at

53:13

UMass. I like to be as specific as possible.

53:17

And part of that is for reasons that we discussed in the talk, is that we have

53:22

patients that are living longer with multiple lesions.

53:25

And it's really tricky for the clinicians when you give them a report that says

53:29

there's three or four lung cancers, and they sort of don't know what to do with

53:34

that. Do they address all of them? Is there one that looks worse than the others?

53:38

Is there one that's going to be potentially life-threatening for the patient versus

53:43

one that we could maybe safely watch that may not become an issue during the rest

53:49

of their expected life?

53:52

And so yes, there's a few really great papers out there that give

53:58

good graphic examples of-- so AAH is atypical adenomatous hyperplasia,

54:04

which is a

54:07

pre-invasive adenocarcinoma spectrum type lesion where it's completely

54:12

ground-glass and it's less than 5 millimeters00.

54:15

So those lesions are often very safe-- not safe, but those lesions are often

54:20

followed because they're small and they're not invasive yet.

54:24

The next step along that spectrum is adenocarcinoma in situ.

54:27

And so this is often characterized by an entirely ground-glass lesion, but it's

54:32

greater than 0.5 centimeters, and they say up to about 3 centimeters.

54:36

So again, this falls in the pre-invasive part of the spectrum.

54:41

It's not an invasive cancer yet

54:44

But as these develop punctate solid components, then they cross into the next part

54:49

of the spectrum, which is minimally invasive adenocarcinoma.

54:53

And this is where management may be indicated.

54:58

And then as you get more and more solid component to your nodule, then those are

55:02

just clearly invasive adenocarcinomas.

55:05

So when I'm reading a case, I will be as specific as possible.

55:09

I'll say, "Slight increase in size of the two-centimeter purely

55:15

ground-glass nodule, which was one centimeter five years ago.

55:20

This likely represents an adenocarcinoma in situ." And I think that when you give

55:24

the clinician that information, then they know how to manage that nodule.

55:28

They say, "Okay, this is still pre-invasive." In an older patient with

55:34

poor lung function and high risk for biopsy, they may say, "We're just going to

55:38

follow this." It might be a different story in a 45-year-old young patient that

55:44

this thing is continuing to grow, and they're going to want to take care of it

55:47

before they're going to have to lose a lot of lung for it.

55:49

So it's very patient-specific, but I think

55:53

we have the information and the data to support that the appearance can really

55:58

reflect the histology. And, of course, it's not perfect.

56:02

Tissue is the issue, but I really encourage people to be as specific as possible

56:07

because it really helps with the management, especially in older patients and

56:10

especially patients with multiple lesions.

56:17

Looks like you cleared through all those questions real quick. Good job.

56:21

Let's see. There was one more here in the chat that I didn't get to.

56:28

Long-standing cysts can transform into malignancies, so for how long should we

56:33

follow? So it's a great question, and it is, I think, still up in the air

56:38

about what the management for these is.

56:40

So Lung-RADS was just updated in 2022, which is the lung cancer screening

56:45

guidelines that we use in the US, and it's the first time that the management for

56:51

the cystic lesions has been incorporated, but it sort of treated them as a

56:56

separate entity. But as we saw in the presentation today, you might

57:02

not just have a cystic nodule or a cyst.

57:06

A cyst could be a component of a complex nodule or mass that has both solid and

57:12

cystic components. So if a nodule has cystic components, do you let the

57:18

solid part of it drive the management, or should there be something separate for

57:22

the cystic component? So I think we need to do a better job at

57:28

characterizing what the cystic part means.

57:31

And, if a nodule is predominantly cystic versus it has minor cystic components,

57:37

are those going to be managed differently?

57:39

So,

57:40

right now we consider a cyst with imperceptible or very thin walls to be

57:47

safe to watch.

57:50

But if there's any sort of wall thickening or nodularity, these are features of

57:55

malignancy that need to be either watched very carefully or sampled and managed.

58:02

And with the cystic ones, PET-CT is often limited because if it's a

58:08

predominantly cystic lesion, we're not really expecting solid tissue to have FDG

58:14

uptake. And so it's like the added value of PET with the

58:20

cystic lesions, either entirely or mostly cystic lesions, is sort of not the

58:26

same as it is for solid lesions. So,

58:30

I'm hoping that the next updates of Fleischner and Lung-RADS will more formally

58:36

address cysts and cystic lesions about how to follow them.

58:41

Let's see. A patient with a known carcinoid tumor in the abdomen and a known SPN.

58:47

Do you prefer dotatate PET or FDG?

58:52

So yeah, this is a specific situation.

58:54

So in the US, often insurance will not reimburse a dotatate PET until you

59:00

actually have tissue. And the reason is because dotatate PET is going to bind to

59:06

those somatostatin receptors, which are overexpressed in the well-differentiated

59:10

tumors. If you have a poorly differentiated tumor,

59:15

they tend to lose their somatostatin receptor expression, and dotatate will be

59:19

negative. So a lot of insurance companies say, "We want to know what type of

59:23

neuroendocrine tumor or carcinoid tumor this is before we go ahead with our PET."

59:29

And so for a well-differentiated, you're going to choose dotatate.

59:32

And for a poorly differentiated, you're going to choose an FDG.

59:34

So if you have a patient with a known carcinoid tumor in the abdomen and they have

59:39

a pulmonary nodule, I think depending on what insurance will allow, either is

59:44

reasonable, but I think most people would probably start with an FDG, especially if

59:49

it's a solid nodule. Because

59:53

for it to be metastatic disease and be singular is not impossible, but a solid

59:58

pulmonary nodule that has suspicious features, I think is probably more likely, if

60:02

you play the odds, to be a lung cancer.

60:06

Let's see. We are at 1:00. There's a couple more questions. I'm not sure, Ben.

60:10

I'm happy to take a couple more, or up to you guys.

60:15

We are at time.

60:21

I will leave that to you, I think.

60:23

Okay.

60:24

Maybe these last couple, and then we'll-

60:28

Okay

60:28

... call it a day.

60:29

How does ILD affect interpretation of lung nodules? So that is interesting.

60:35

I think formally, in terms of the guidelines, there really isn't anything that

60:40

says, oh, you would manage a nodule differently in this patient who has ILD versus

60:45

this patient that doesn't. But personally, in my practice, I think the presence of

60:50

ILD makes me think outside the box a little bit more.

60:53

If there's increasing solid tissue, but it's not really a nodule, and it's more

60:58

sort of sheet-like or non-mass-like, I think to myself, could this be a cancer?

61:04

And does it just not look typical because the background that it's growing on is so

61:10

abnormal?

61:12

So personally, for me, I tend to think outside the box a little bit more and will

61:18

include things that in a patient with normal lung, I would say, "Hmm, that's

61:23

probably scarring or atelectasis." But in a patient with ILD, I might be a little

61:27

bit more suspicious about it. Ask for follow-up or recommend a PET or something

61:32

like that.

61:34

Okay. And here's the last one. Do you see a large ground-glass opacity about

61:40

three centimeters being stable from more to three to five years or longer in your

61:43

experience? Yeah. So a lot of the guidelines sort of talk about following things

61:49

out to two or three years. But

61:53

in my practice, I have definitely seen things that change or that are stable for

61:59

three or four years. And then, if you go back and you see something from 10 years

62:04

ago, then you say, "Oh, there definitely has been a change here." So I think

62:10

that there are very, very slow-growing adenocarcinoma spectrum type lesions.

62:16

They usually start as an AH or an AIS that

62:22

the guidelines say, oh, you can stop following after two to three years.

62:26

But,

62:27

I will say in my personal practice, I've definitely seen ones that have been

62:31

followed for longer, or we have an old comparison, and we do see changes after that

62:37

time. So stability of two to three years makes you feel better, but may not be

62:42

enough. So I think the more we learn about lung cancer screening and the longer

62:47

we're tracking these lesions, we'll learn more about that.

62:51

So,

62:53

yes, I hope that answered your question, but thank you guys so much for your

62:58

attention and interest, and these were some really great and interesting questions

63:03

that I'm going to add some things to my

63:06

lecture to address some of these. So thank you.

63:10

Yeah. Dr. McIntosh, thanks for that lecture and for answering all those great

63:13

questions. And thanks to everyone who joined us today.

63:17

Okay.

63:18

Be sure to join us next week on Thursday, October 22nd at 12:00 PM Eastern, where

63:23

Dr. Steven Pomeranz will deliver a lecture entitled, "Deep Dive into Foot and Ankle

63:29

MRI." You can register for that at medallity.com and follow us on social media for

63:33

updates on future noon conferences. Thanks again, and have a great day.

Report

Faculty

Lacey McIntosh, MPH, DO

Director, Oncologic Imaging; Assistant Professor, Radiology

University of Massachusetts Medical School / Memorial Health Care

Tags

Chest