The potential of shear wave elastography — an update

First and foremost, we would like to thank Prof. Ghulam Nabi of the University of Dundee for providing us with a preprint of the full text of his team’s article on transrectal SWE within about 30 seconds of receiving our request! That’s a record!

Having now had the chance to read through the entire article by Wei et al. (as opposed to just the abstract), we are in a much better position to add some comments as to the potential of transrectal SWE that reflect the limitations of the study reported in this article.

As reported in our initial commentary, Wei et al. state in the abstract of the paper that “Potential limitations include selection bias and study being single centre site.” Both of these limitations are important, but the issue of selection bias is extremely important.

What was actually done in this study, over a 45-month period from November 2013 to August 2017, was as follows:

  • The research team gave transrectal SWE tests to 218 men who
    • Had already been diagnosed with prostate cancer and
    • Had already selected laparoscopy radical prostatectomy as their treatment and
    • Were actually treated by laparoscopic radical prostatectomy immediately after they were given a transrectal SWE

As a consequence, the types of patient enrolled in the study were extremely likely to have higher-risk forms of prostate cancer, as is clearly reported by Wei et al.:

  • 212/218 patients were eligible for evaluation (i.e., six initial enrollees were excluded for various reasons)
  • The average (mean) age of the patients evaluated was 67.6 ± 5.4 years
  • The average (mean) PSA level of these patients at diagnosis was 11.8 ± 8.1 ng/ml (range, 0.9 to 47 ng/ml)
  • The average (mean) prostate volume of these patients was 66.9 ± 30.4 ml (range, 20 to 207 ml)
  • The pathological Gleason scores of the patients (post-surgery) were
    • 3 + 3 = 6 in just 7 patients (3 percent)
    • 3 + 4 = 7 in 101 patients (48 percent)
    • 4 + 3 = 7 in 35 patients (17 percent)
    • 8 or higher in 69 patients (31 percent)
  • The pathological stages of the patients (again, post-surgery) were
    • pT2a in 11 patients (5 percent)
    • pT2b in 3 patients (1 percent)
    • pT2c in 106 patients (50 percent)
    • pT3 or pT4 in 25 patients (12 percent)
    • pN0 in 187 patients (88 percent)
    • pNx in 12 patients (6 percent)
    • pN1 in 13 patients (6 percent)
  • 120 patients (57 percent had positive surgical margins); the other 92 (43 percent had negative surgical margins)

In other words, based on their Gleason scores alone,

  • 205/212 patients (97 percent) had either intermediate-risk or high-risk disease.
  • 104/212 patients (49 percent) had either unfavorable intermediate-risk or high-risk disease.

So, we should be clear that the article by Wei et al. is not an article that was about testing the accuracy of SWE in the diagnosis of prostate cancer. It is an article about a study investigating the accuracy of SWE as a test that could assess, prior to surgery, the actual pathological findings post-surgery in a group of men who almost all had intermediate- or high-risk prostate cancer. And this is a very valid form of study, but it does need to be seen in appropriate context.

So given that context, what did Wei et al. find?

They were able to show that, when the SWE findings (as assessed by a trained expert) were used to predict the actual histological findings post-surgery (as assessed by a skilled uropathologist), the SWE data were able to predict the actual histological findings with a surprisingly high degree of accuracy:

  • Sensitivity was 88.6 percent
  • Specificity was 97.3 percent
  • Positive predictive value (PPV) was 86.3 percent
  • Negative predictive value (NPV) was 97.8 percent

These are impressive numbers. Do they mean that transrectal SWE can be used to diagnose prostate cancer?

No. What they do is imply that this is a possibility.

At a minimum, one large, multi-center study is going to need to be carried out in which a cohort of men who are suspected to be at risk of localized prostate cancer are initially evaluated in some way similar to the following:

  • All given a transrectal SWE, which is used to predict their risk for prostate cancer of differing types (or not) by experts in reading transrectal SWE images and who are allowed to know only the patients’ PSA levels, DRE results, and other relevant symptomatology at the time they make this assessment and
  • All actually diagnosed with prostate cancer (or not) based on traditional methods (PSA level, DRE results, TRUS or other imaging such as MRI scans, and a biopsy) with no reference to the transrectal SWE data

The hard question will be to work out what is the appropriate management / treatment for the men in this study — especially when the transrectal SWE suggests more aggressive  or less aggressive management than the traditional methodology. We’ll leave that to specialists in clinical trials to work out, but there are serious ethical issues here that will need to be resolved.

With regard to potential endpoints for such a study, these might include some or all of the following:

  • Can SWE determine with accuracy which patients don’t need a biopsy at all (and based on what criteria)?
  • Can SWE determine which patients can be appropriately followed on active surveillance as well as or better than traditional methods (based on predetermined criteria for active surveillance candidates)?
  • Can SWE determine which patients absolutely need immediate treatment better than traditional methods?
  • What level of accuracy did SWE show when compared to histopathological findings in men who elected to have a radical prostatectomy as their form of treatment?
  • What were levels of disease progression (in men on active surveillance) and levels of biochemical recurrence (in men who received treatment) and how well did these correlate to predictions based on SWE and on various forms of treatment nomogram?

And of course every center involved in such a study will need to have had sufficient experience in the  conduct and interpretation of SWE — probably with a central review capability for all SWE data as well.

So what we have here is a potential series of hypotheses that:

  • Transrectal SWE may be able to eliminate the need for immediate initial biopsies in some men suspected to be at risk for prostate cancer based on PSA and other data.
  • Transrectal SWE may, together with PSA and other data, be used to monitor the need for biopsies over time in some patients.
  • Transrectal SWE may, on its own or in association with biopsy data when required, be able to simplify the diagnosis of prostate cancer.
  • Transrectal SRE may, on its own or in association with PSA and other data, be useable to monitor men on active surveillance, eliminating the need for  expensive MRIs

This is definitely an interesting possibility … but your sitemaster is still not convinced (yet) that it will turn out to be more cost-effective than current methods.

The paper does not give us any precise descriptors of the way in which transrectal SWE was carried out in these 218 patients, but it does state that the SWE was carried out using

an endocavity Aixplorer® ultrasound transducer (SuperSonic Imagine, Aix-en-Provence, France) through the rectal wall focusing on the prostate avoiding any pressure on the transducer.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: