Projecting the need for confirmatory biopsies in candidates for active surveillance

A newly published paper from the Memorial Sloan-Kettering Cancer Center group in New York has proposed a predictive model which might allow a high percentage of candidates for active surveillance as initial management to skip a second, confirmatory biopsy after initial diagnosis.

Most protocols currently recommend that patients being considered as candidates for active surveillance after initial diagnosis should be given a second, confirmatory biopsy — ideally under MRI/TRUS fusion-guidance biopsy — before the patient is considered to have low-risk disease suitable for initial management by prospective monitoring. However, this new paper by Satasivam et al., in the January issue of the Journal of Urology, now suggests that at least 75 percent of those patients may, in fact, be able to avoid that confirmatory biopsy. This is likely to be seen by many as a risky idea, but the data provided by the authors is really rather compelling (although they do state very clearly that their model needs external validation before being accepted).

Satasivam et al. evaluated data from a cohort of 392 patients, all initially diagnosed with Gleason 6 disease, all initially diagnosed between December 2007 and December 2013, and all of whom actually underwent a confirmatory biopsy. They set out to determine whether data from the initial biopsy could be used to predict the results of the confirmatory biopsies for these patients, and whether, therefore, some of these patients could actually skip the confirmatory biopsy altogether.

Here is what they learned from their study:

  • The average (median) age of their patients was 62 years (range, 56 to 66 years).
  • MRI scans were carried out on all patients at a median of 3 months after their initial biopsy and diagnosis.
  • Confirmatory biopsies were carried out on all patients at a median of 5 months after their initial diagnosis (but none of these men received an MRI/TRUS fusion-guided rebiopsy because this type of biopsy was largely unavailable at that time).
  • 184/392 patients (47 percent) had a dominant/focal lesion on MRI, and 208 (53 percent) did not.
  • On confirmatory biopsy
    • 135/392 men (34 percent) had no cancer visible.
    • 213/392 men (54 percent) again exhibited Gleason 3 + 3 = 6 disease.
    • 39/392 men (10 percent) were upgraded to Gleason 3 + 4 = 7 disease.
    • 1/392 men (0.3 percent) was upgraded to Gleason 4 + 3 = 7 disease.
    • 3/392 men (0.8 percent) were upgraded to Gleason 4 + 4 = 8 disease.
    • 1/392 men (0.3 percent) was upgraded to Gleason 9 disease.

The authors were also able to show that a predictive model, based on multivariate analysis and using only the total tumor length in the initial, diagnostic biopsy cores along with the patient’s PSA density was significantly associated with the actual occurrence of a Gleason score of 7 or higher on confirmatory biopsy, with an AUC of 0.85. They also suggest that use of this predictive model would be able to predict the patients who truly needed a confirmatory biopsy with an accuracy of down to 97 percent percent. In other words, it would only miss an upgraded cancer if the urologist felt obliged to carry out 50 or more confirmatory biopsies to find a single case of upgrading.

The authors also explain that, if the doctor and patient are only willing to carry out a confirmatory biopsy if the model demonstrates a 10 percent risk for upgrading on rebiopsy, then

  • The model would project the need for confirmatory biopsies in 240/1,000 patients.
  • This would reduce the need for confirmatory biopsies among candidates for active surveillance bu 75 percent.
  • Of the 240 patients who would undergo rebiopsy based on these criteria, 89 (37 percent) would be found to have a cancer with a Gleason score of 3 + 4 = 7 or higher.
  • Of the 760 patients who would not undergo rebiopsy based on these criteria
    • 20 (2.6 percent) would actually have cancer with a Gleason score of 3 + 4 = 7.
    • 3 (0.4 percent) would actually have cancer with a Gleason score of 4 + 4 = 8.

If the accuracy of this model can indeed be confirmed through external validation by an independent team of investigators (which should not really prove difficult), this would allow doctors — rather than simply telling their patients that a confirmatory rebiopsy is an essential component of entry onto an active surveillance protocol — to use this model to either tell patients that they need a rebiopsy (because they are at significant risk for an upgrade) or have a conversation with their patients about whether the patient would like to have a rebiopsy or skip it, given the low risks of an upgrade in their case.

It is interesting that the model proposed by Satasivam et al. does not take any note of the possibility of genomic/genetic testing to evaluate risk for aggressive disease. It might be interesting to know whether use of such testing could improve the predictive accuracy of this model or actually made it less accurate. If it made it less accurate, this would raise questions about the utility of the currently available genetic/genomic tests.

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 )

Google photo

You are commenting using your Google 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: