MRI Online
Library

View this list on the Library Page

Pricing
MRI Online
MRI Online
MRI Online
Library

View this list on the Library Page

Pricing

PI-RADS 4 Initially Read as PI-RADS 3

6 min.
Pomeranz, Stephen
Stephen J Pomeranz, MD
Chief Medical Officer, ProScan Imaging. Founder, MRI Online
CME Eligible
Includes DICOM files

Dr. P. here with a case in which the reading was a PI-RADS 3 and should have been a PI-RADS 4. It's a 68-year-old man. His PSA four months ago was 5.87. One year ago he had a biopsy which was scored as a Gleason 7 or a 3+4 equals 7. And we have his b1600 diffusion image which is a little iffy but still diagnostic. Now, remember from our prior vignettes that I like to have a b-value of say 50, a b-value of 700 or close to it, 600, 700, 800, and I use these to make my ADC map and I keep my high b-value image. In other words, 1400, 1500, 1600 separate, to look at the focal area of diffusion restriction as a nice bright white area.

Now, in this case they didn't obtain the ideal lower b-values, but that's not really what I'm here to discuss. But I will say that when you have a hampered diffusion series, it should be given the designation category X.

In breast imaging, we often use the term maybe BI-RADS 0 to indicate that additional imaging is needed, but category X is used to define the diffusion series that is hampered. And in that scenario, you have to rely heavily on the T2 and the dynamic contrast enhanced MRI or DCE MRI.

As we scroll through this case, there's an area in the left PZ, including the PZpm, postero-medial and somewhat posterior and maybe even a little postero-lateral, and it's on two or three consecutive cuts. And on the contralateral side, there's a second area that stands out on the b1600 diffusion image. Now, when we look on the ADC map in the center, that bright area is pretty dark. And similarly on the right side, the area that was pretty bright, not so congruent, so this area is a little bit harder to define, but it's at least an area where the biopsy should be focused. Right here, I think there is concordancy between the ADC map and the b1600 diffusion image. So those are going to be two extra areas that are biopsied, at the time of intraoperative ultrasound with MR correlation and guidance.

So what happened? The patient had 15 cores. Of those 15 cores, 5 were positive. They were all Gleason 7s. They were all 4+3 Gleason 7s. And in fact, this area, as well as this area, which has a correlate on the ADC map, we're both Gleason 7s. So both sides of the gland were involved, making it a T2-C gland. In other words, both lobes, a T2-A gland would be less than 50% of the volume on one side. A T2-B gland would be greater than 50% of the gland on one side. A T2-C gland would be both sides of the gland or both lobes of the gland, and that's the case here, and that's how it was read by the pathologist. In addition of each core, the highest percent involvement of a core was 26%, which is a pretty good volume. There was no evidence of neurovascular bundle invasion. There was no evidence of capsular invasion. There were no signs of nodal involvement and the patient underwent a prostatectomy.

Now, before we talk about the prostatectomy, let's look at this under-called study. Here's the axial T2 weighted image. Let's look very carefully. If we T2 assess this, just based on the T2 characteristics alone in the peripheral zone, I think you would all admit that this is not a uniform hyperintense area, so a score of 1 is out. It's not linear or wedge-shaped, so a score of 2 for just the T2 character in the peripheral zone is out. A score of 3, heterogeneous, ill-defined, non-circumscribed, some areas rounded, moderate hypointensity. I think you could accept that as a possibility. A score of 4, circumscribed, homogeneous, moderately hypointense, focal with mass effect, confined to the prostate, and less than 1.5 centimeters. Well, I'm not so sure about that. I think all of those statements could be true except confined to the prostate, and this is where I think the reader made an error.

Look at the right side, a crisp hyperintense peripheral zone on the right. You can see the periprosthetic fat. But on this side, not so much, a little globular, a little gray, a little ill-defined, definitely asymmetric. Bright, not so much. There should have been an inference that there was capsular abutment and a suspicion of neurovascular bundle spread. And on prostatectomy, Gleason 7 was confirmed. 10% of the entire gland was involved by Gleason 7. There was capsular involvement, there was left-sided neurovascular bundle involvement.

Now, let's go back to the diffusion image and make one more point that has been made in PI-RADS 2.1. And that is, when you have bilaterality, if it's asymmetric, that's a concern. Symmetry, just like in pediatrics and anywhere else in radiology is usually a sign of benignity, but this is not symmetry. Look at this lesion and now look at this lesion. They're shaped differently, they're positioned differently, and you can confirm that or corroborate that on the ADC map. Mid to right-sided lesion, left-sided lesion, in no way are they symmetric.

So those are some critical pitfalls. This one was under-read. There was capsular invasion. There was neurovascular bundle invasion. It should have been at the very least, read as a PI-RADS 4. It is a corroborated Gleason 7.

LESSON 3, TOPIC 5

Mastery Series: PI-RADS 2.1 Update

Mastery Series

Browse other topics in...

Content reviewed: December 29, 2021

View CME activity information