A compound genetic/genomic and microenvironmental risk classification for localized prostate cancer


As we become more knowledgeable about the underlying genetic and other factors that affect the risk for, and the risk for progression of, apparently localized prostate cancer, we also become more able to predict such risks … but this is a very complex topic. A new paper by Lalonde et al. in The Lancet Oncology (abstract or full text article) has given us a clear picture of just how complex this issue may be. The paper itself is supported by a set of editorial comments (“Chromoplexy and hypoxic microenvironment drives prostate cancer“) by Ballas et al.

What Lalonde et al. have shown — for the first time — is that risk for progression of apparently localized prostate cancer after first-line therapy can be projected based on an integrated combination of genetic (DNA-based) and tumor hypoxic (microenvironment-based) indices that appear to characterize the probability of failure of first-line therapy.

As the authors note, we have long known that some 30 to 50 percent of patients who receive standard first-line treatments (e.g., radical prostatectomy or image-guided radiation therapy) have at least a biochemical recurrence (even if no apparent clinical recurrence). They therefore set out to see whether they could find new and better ways to project which of these patients were at elevated risk for failure of such first-line therapy. To do this, they decided to look at both

  • Tumor DNA-based data alone and
  • Tumor DNA-based data and measurements of intra-prostatic hypoxia

They used data from a series of 126 low- to intermediate-risk patients (all scheduled for first-line therapy with IGRT) to develop four different prognostic indices. They then used clinical data and pathological data from the prostatectomy specimens of two independent cohorts of 154 and 117 radical prostatectomy patients who had been treated for low- to high-risk prostate cancer to validate the four different prognostic indices.

The primary endpoint for the study was a set of prognostic measures that could be used with high accuracy to stratify patients into categories for risk of biochemical relapse with 5 years after their primary treatment. Here are the key findings of the study:

  • Based on multivariate analyses, biochemical relapse was associated with indices of
    • Tumor hypoxia
    • Genomic instability
    • Genomic subtypes
  • Four genomic subtypes of prostate cancer were identified, each of which had a different probability for biochemical relapse-free survival at 5 years.
  • Genomic instability was found to be prognostic for relapse in patients treated by
    • IGRT (hazard ratio [HR] = 4.5; p = 0.00013)
    • Radical prostatectomy (HR = 4.0; p = 0.0024)
  • The impact of genomic instability was magnified by intra-tumoral hypoxia (HR = 3.8; p = 0.019).
  • A novel 100-locus DNA signature accurately classified treatment outcome in the 126-patient low- to intermediate-risk
    cohort (HR = 6.1; p = 0.0015).
  • In the independent 154- and 117-patient radical prostatectomy cohorts, this novel 100-locus DNA signature
    • Identified low- to high-risk patients who were most likely to fail treatment within 18 months (HR = 2.9; p = 0.0039)
    • Was better at predicting biochemical relapse than 23 previously published RNA signatures.

The authors conclude that patients who exhibit these aggressive features after initial biopsy should not receive simple, standard, first-line therapy, but need to be be entered into “treatment intensification” trials, i.e. trials of more aggressive forms of first-line and combination therapies. However, carrying out prognostic profiling of this type on all prostate cancer patients would still be a massive challenge today.

In their editorial comments on this paper, Ballas et al. note that:

  • Strictly speaking, Lalonde et al.’s findings are “hypothesis-generating”, and will need to be validated prospectively in other studies, but …
  • These findings may well have therapeutic implications (because some mutations can be targeted therapeutically and it may be possible to modulate tumor hypoxia too).
  • A lot of work will be needed to work out “the best combination of targeted therapy for the cancer cell and modification of the microenvironment”.
  • Even so, all that such modification can promise at present is improvements to standard forms of first-line therapy.

Major progress in the first-line management of localized prostate cancer may be possible if we can break localized adenocarcinoma of the prostate (by far the most common category of prostate cancer) down into a series of much better defined subcategories, based on genetic, genomic, and other factors (e.g., microenvironmental ones, like tumor hypoxia levels). We aren’t there yet, but papers like this one do suggest that we are getting closer. If these subcategories can be validated, the next question is going to be whether and how to standardize such classification of prostate cancers on a routine clinical basis.

One Response

  1. This sounds much like some of the current commercial genetic tests, like Decipher, that I believe is offered post-RP.

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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: