0:01

So here's our typical UIP

0:02

or sometimes people will still call it UIP pattern.

0:04

And so this is our highest competence pattern of UIP.

0:08

And so what are we looking for in HR ct?

0:11

You wanna see peripheral basal predominant

0:13

pulmonary fibrosis.

0:14

So you're gonna see reticulation,

0:15

these small little lines superimposed on each other

0:19

with associated subpleural honeycombing.

0:21

So these fibrotic cysts which are lining up in rows and

0:25

or stacking upon each other within the most peripheral

0:28

portion of the lung, you can have associated traction

0:31

bronchiectasis, but you don't need traction bronchiectasis

0:34

to define this pattern.

0:36

The last part's very important.

0:37

You wanna have absence of features

0:39

that would be non-consistent with UIP.

0:42

So things that that would be pushing you away from

0:45

UIP diagnostic pattern.

0:46

Things like significant air trapping,

0:48

diffuse nodule lung disease or abnormal distributions.

0:53

And here's the classic pattern of UIP here.

0:56

I think a medical student probably could make

0:58

this diagnosis quite easily.

0:59

Uh, peripheral basal predominant

1:01

with these sequential axial images.

1:03

Clearly honeycombing.

1:05

Look at these beautiful honeycomb cysts lining up in rows

1:07

and clearly stacking upon each other.

1:10

No other findings that suggest alternative diagnosis.

1:12

This is gonna be UIP.

1:15

Another example of UIP here.

1:17

I think the coronal reformations are really helpful

1:19

for these HRCT scans.

1:21

In terms of distribution,

1:23

we can clearly tell this coronal reformation

1:25

that there is a basal gradient.

1:27

The lung bases are more severely affected than the

1:29

upper aspect of lungs.

1:31

And this looks to be a peripheral disease process.

1:33

So on one coronal slice we can be pretty confident.

1:36

This is a peripheral basal predominant process

1:39

and we see plenty of subpleural honeycomb.

1:41

Let's blow this up so you can see it.

1:42

Beautiful honeycomb cyst lining up in rows,

1:45

stack you upon each other.

1:47

Take you to the bank. This is going to be UIP.

1:50

Okay, another example here, UIP.

1:53

This is less floored case,

1:55

but on this axial slice we clearly see

1:57

that it's peripheral lung preponderance.

1:59

Pull up the coronal

2:01

and again, coronal is so great for distributions.

2:03

Clearly basal predominant looks, peripheral predominant.

2:05

And you see these small subpleural cysts.

2:09

The beauty of this high confidence UIP pattern is

2:13

that you've essentially obviated sub surgical lung biopsy.

2:16

You just don't need to do it.

2:18

So at least 95% of the time,

2:21

if you have this pattern on HRCT,

2:24

you're gonna get UIPN pathology.

2:26

That's pretty much as high

2:27

as you can get in terms of any sort of test.

2:29

Very, very high. So in these cases, we seldom,

2:32

if ever we'll even think about getting surgical lung biopsy.

2:36

And so this is quite, quite helpful.

2:37

We almost now will equate this pattern

2:40

to a UIP pattern on pathology.

2:45

Now let's talk about probable UIP.

2:47

And so the beauty of probable UIP is that if you know

2:51

what the high confidence typical UIP pattern looks like on

2:54

ct, you know a probable UIP pattern looks like on

2:57

HRCT. And so

2:58

It's peripheral base and predominant.

3:01

And in this setting you also wanna see some traction

3:04

bronchiectasis or bronchiectasis.

3:07

'cause the main difference between this probable UIP pattern

3:10

and the more high competence UIP pattern is

3:12

that there's no honeycombing in probable UIP.

3:15

So because there's no honeycombing,

3:17

we wanna see other findings of definitive fibrosis.

3:20

And that's gonna be the traction bronchiectasis

3:22

or bronchiectasis telling us

3:24

that indeed this isn't just mild reticulation,

3:26

mild ground LACS opacity.

3:28

There is real fibrosis within that lung parenchyma.

3:31

And the most high confidence way to tell

3:33

that is if there's either honeycombing and

3:36

or traction bronchiectasis.

3:37

So in this disease setting, we wanna see some traction,

3:39

bronchiectasis or bronchiectasis.

3:42

And again, nothing else that draws you away from

3:44

a diagnosis of UIP.

3:45

So things like air trapping,

3:46

nodule lung disease, abnormal distributions.

3:50

So a lot of these probable UIP cases, bottom line,

3:53

they look like just less severe cases

3:56

of the more floored UIP.

3:58

And so here's an example here.

4:00

Sequential axial images, clearly basal predominance,

4:02

clearly peripheral and preponder as well.

4:05

Some mild subpleural traction.

4:08

Bronchiectasis and bronchiectasis, no honeycombing though.

4:12

So if you look carefully, there's no

4:14

fibrotic cysts which are lining up in rows

4:16

or stack upon each other.

4:18

So problem UIP.

4:20

So I talked about the typical UIP pattern

4:22

and a strong correlation with pathology.

4:24

Let's talk about the problem UIP pattern.

4:26

So this used to be contentious.

4:28

I don't think it's contentious anymore.

4:29

So as I alluded to

4:30

before, if you have this high confidence,

4:32

typical UIP pattern, take it to the bank,

4:34

it's gonna be UIPN pathology, the probably UIP pattern.

4:38

This took some, some work,

4:40

but eventually we were able to figure out that the yield

4:42

of UIPN pathology was also very high.

4:45

Somewhere in the 80 to 90% range,

4:48

but not as high as in typical UIP.

4:51

But what this allows us

4:52

to do is actually dig a little deeper

4:54

because we have this data

4:55

and we, we also found

4:56

that if you have a high pretest probability

5:00

of idiopathic pulmonary fibrosis,

5:02

idiopathic pulmonary fibrosis is a disease of older people.

5:05

So the patient is older.

5:07

See older than 60

5:08

or 65, no known cause for interstitial lung disease, right?

5:12

Because again, idiopathic pulmonary fibrosis is

5:14

an idiopathic disease.

5:16

No other risk factors, no

5:17

inhalation exposures, nothing like that.

5:18

So they're coming in and even before, even

5:21

before the patient comes in,

5:24

people are thinking this patient has IPF,

5:25

all you need is a probable UIP pattern to define that IPF

5:31

clinical diagnosis.

5:32

So in that specific disease setting,

5:34

probable UIP is quite powerful.

5:36

And I say at least in clinical practice, the majority

5:39

of our patients who come in with pulmonary fibrosis

5:41

as a new diagnosis, again, most

5:43

of these patients are gonna be IPF

5:45

or at least as a plurality.

5:46

And so very seldom with these probably UIP patterns,

5:50

are we pursuing surgical lung biopsy also, again,

5:53

'cause of that concomitant risk

5:54

of surgical lung biopsy in patients with pulmonary fibrosis.