Targeting bone metastases with radiation in oligorecurrent men has no survival benefit in Mayo study

Oligometastases in Bones

Metastasis-directed therapy (MDT) when there are only a few bone metastases (called “oligometastatic”) is controversial. It can certainly relieve pain, and prevent fractures and spinal compression. It can also provide good “local control” (cancer in the irradiated metastasis is permanently destroyed) and reduce the PSA that those metastases put out. But is there any survival benefit?

Patients often ask radiation oncologists (ROs) for radiation of those metastases using targeted radiation (which I’ll call “zapping”), and they ask their ROs to treat new metastases as they are detected. This is called “metachronous treatment,” but I’ll call it “whack-a-mole” Sometimes metastases appear in places where radiation treatment may be problematic, such as near vital organs or deep in the spine. The nagging question is whether such treatment really does the patient any good. With the approval of ever more sensitive PET scans, like the PSMA PET scan approved last week, patients will undoubtedly detect more metastases.

The Mayo Clinic has been one of the cheerleaders for MDT. They have posted a somewhat misleading YouTube video featuring their C-11 choline PET scans, but showing only how good the local control is. What the video can’t show is how those patients would have done without MDT — there was no control group ever used or shown in their video.

Perhaps to partially correct for the misleading video, Boeri et al. at Mayo retrospectively looked at 115 patients who had an oligometastatic recurrence to the bones (1 to 5 metastases):

  • 115 patients were treated with SBRT. They had a median of 1 bone metastasis.
  • 47 patients were treated with ADT only. They had a median of 2 bone metastases.

This was not a randomized study, so it is entirely likely that there was “selection bias” — for example, patients who received ADT only may have done so because it was felt they would not be able to benefit from SBRT or that it might be unsafe. Patients who received ADT only had a higher number of bone metastases and a higher PSA. All of those receiving MDT for bone metastases were also receiving ADT.

  • The 5-year prostate cancer mortality was no different between the two groups
  • The 5-year radiographic recurrence-free survival was no different between the two groups
  • Among those with 5 years of follow-up, the time remaining free of the next significant systemic therapy (e.g., chemotherapy, Zytiga, etc.) was longer for those getting zapped. However, it should be noted that the decision to give an additional significant therapy is a physician decision based on many factors, including patient status, number of metastases, and PSA. Because number of metastases and PSA are changed by MDT, and those receiving MDT started with one less metastasis, the physician may feel pressured to start a new therapy sooner in patients receiving ADT only.

Pending confirmation from long-term randomized clinical trials of MDT to oligometastases in bones, there is no evidence of oncological benefit.

Oligometastases in Pelvic Lymph Nodes (PLNs)

MDT of oligorecurrent metastases that are evident only in pelvic lymph nodes (PLNs) is less controversial. Lymph is a slow-moving fluid, and metastatic cancer cells emerging from the prostate might get trapped in the lymph nodes that drain the prostate. So it has been hypothesized that treatment of the PLNs when a few are found to be cancerous may still provide a cure. This has not yet been proven in a randomized clinical trial, but there is observational evidence of a significant benefit to salvage whole-pelvic radiation (see this link).

What is controversial about the way they are treated at the Mayo Clinic is that only those cancerous PLNs and a small margin around them were surgically removed, and whole pelvic salvage radiation wasn’t routinely given. They were treated in any of three ways:

  1. Salvage pelvic lymph node dissection (sPLND). Jeffrey Karnes at Mayo is one of the few top surgeons in the US who does this difficult surgery. It is difficult because PLNs detected on a PET scan can be very small. They can also be invisible, hidden in fat deposits, and very difficult to find. There are innovative techniques like fluorescent or gamma-ray PSMA indicators that can facilitate detection. Patients treated with sPLND also received 6 weeks of bicalutamide. This group actually included some patients who had lymph nodes removed that were outside the pelvis (see below).
  2. External beam radiotherapy (EBRT) to PLNs as part of salvage radiation treatment (sRT). At Mayo, 72 percent received salvage intensity-modulated radiation therapy (IMRT) to the identified PLNs plus a large margin around them, while 28 percent received stereotactic body radiation therapy (SBRT) to just the identified PLNs plus a small margin around them. This was typically done along with 12 to 18 months of ADT.
  3. ADT only, Patients treated with either of these two forms of MDT were compared to patients who received ADT only, which is the current standard-of-care. Again, this was not part of a randomized clinical trial, so it is likely that the ADT only patients were not offered MDT for a reason. Most importantly, about half had cancerous lymph nodes in the retroperitoneum or abdomen (stage M1a) — already outside of the prostate drainage area (stage N1), and they had more positive lymph nodes. In contrast, only 9 percent of the sPLND group and 19 percent of the EBRT group were found to have cancerous lymph nodes outside the pelvis. The ADT only group had much further progression at the time of treatment.

After a median follow-up of 47 months:

  • Prostate cancer-specific mortality was 13.5 percent for ADT only, 9.5 percent for EBRT, and 6.3 percent for sPLND (difference between ADT only and sPLND was statistically significant)
  • Radiographic recurrence was 65 percent for ADT only, 40 percent for EBRT, and 61 percent for sPLND.
  • Castration-resistance was 39 percent for ADT only, 19 percent for EBRT, and 21 percent for sPLND.
    • The median time until castration resistance set in was 59 months for ADT only, 73 months for EBRT, and 98 months for sPLND.
  • Second-line systemic therapies were offered to 43 percent for ADT only, 29 percent for EBRT, and 24 percent for sPLND.
    • The median time until the therapies were offered was 28 months for ADT only, 32 months for EBRT, and 44 months for sLND.
  • Inexplicably, the percentage of cancerous lymph nodes outside of the pelvis (% M1a) was not included as a variable to correct for in their multivariable analysis, and was largely ignored.

The authors found an association between MDT and radiographic progression in their retrospective sample of patients. However, it leaves unanalyzed how much of that association is due to the extraordinarily high rate of out-of-pelvis progression already present in the patients treated with ADT only. In fact, it seems likely that that is the reason they didn’t receive MDT.

They also make the same error with respect to onset of castration-resistance and use of second-line therapies that they made in their bone MDT analysis; i.e., they “treated PSA” with their MDT, so they can’t use onset of castration resistance and time to second-line therapy as useful endpoints. Tellingly, radiographic recurrence is similar for the ADT only and sPLND groups of patients, while being lower for the EBRT group, possibly only because of the longer use of adjuvant ADT with EBRT.

Another open question is whether whole pelvic salvage radiation might have been more effective than the limited margins they used at Mayo. With the more accurate PSMA PET scans, ROs are able to treat the entire PLN area with radiation boosts given to the detected ones. The RTOG consensus treatment area has recently been expanded (see this link). It’s important that patients understand the detection limits of even the best PSMA PET scan: metastases smaller than 4 mm, and those that put out only small amounts of PSA remain invisible.

Editorial note: This commentary was written by Allen Edel for The “New” Prostate Cancer InfoLink.

7 Responses

  1. Since SBRT is quick (only three fractions per lesion) and free, I see no reason not to zap visible oligomets. I’ve done it to the three bone lesions, one in 2017 (detected by Axumin) and two in 2018 detected by Ga-68 PSMA. In addition, I switched to estradiol patches in early 2019 and have had no recurrence since and am still castrate sensitive (serum T = 3.0 ng/dl). I’m comforted by the combination therapy.

  2. Allen,

    Excellent discussion — thank you for your analysis. As a patient at Mayo in 2013 for a sLND, I have had time to ponder whether the “whack-a -mole” approach is worth pursuing. As you mention, Mayo added 6 weeks of bicalutamide after the sLND, which in my opinion, tainted the results and comparison analysis. Further difficulties are added to the surgery if the patient previously had RT to the surgical field, and there may be a disconnect between what the choline PET scan shows as positive or negative nodes, and which nodes the surgeon actually removes. Some results of sLNDs at Mayo can be found here:

    A realistic goal of MDT would be to delay systemic therapy, or alleviate pain if that is the problem. Achieving a cure and extending overall survival, while possible, are probably not realistic goals, given the current state of technology. If the patient’s goal is a material delay in systemic therapy, to set the goal post, the patient should first determine how long of a delay is “material”. Then the patient should carefully evaluate the characteristics of his own prostate cancer (Gleason score, PSA level, PSA doubling time, scan results, life expectancy, etc.) and factor in potential morbidity of surgery and cost of the procedure.

    If the patient has a fast (i.e., short) PSA doubling time, any delay in systemic therapy will be short lived.

    Best regards,


  3. Richard:

    Slowing progression would be an admirable goal. Unfortunately, there is no good evidence yet that MDT accomplishes that.

  4. Allen:

    What is your view on whether there is an abscopal effect and if it can be studied?

    We have seen a couple of highly reputable genitourinary med/oncs acknowledge the abscopal effect recently; and we have seen one man, of course anecdotally, see his PSA fall post spot SBRT before indeed it did rise. AnCan has long been aware of this effect — some of us remember “Lupron Jim” GRHS who died from a cancer other than prostate.

    As a rule, AnCan moderators see spot RT (SBRT/IMRT) only as an alternative for palliative Mx pain — not systemic. And we agree with you that Kwon at Mayo frequently turns to spot RT and surgical removal of Nx disease more frequently than others.

    It is interesting to AnCan how the Mayo docs spin these study results to their patients and the outside world!

  5. Hi Rick.

    I knew you’d be interested in this — I know we both have always been skeptical about the claims made for MDT and the way they do it at Mayo.

    Yes, as you may know, I was the one who told Lupron Jim (z”l) about the abscopal/bystander effect and went with him to Dr. King to discuss it. Jim had Provenge and LN RT. I am a true believer in it.

    To get a large and sustained response requires some kind of immune amplification, as was used with Xofigo + Provenge or metastasis-directed SBRT + ipilimumab in the two studies we looked at recently.

    Prostate cancer is just too immunologically “cold” to get a sustained abscopal/bystander response without some kind of boost. Perhaps if Mayo had used some immunogenic adjuvant therapy along with MDT they might have gotten a better response. Unfortunately, Provenge is not approved for men who are recurrent but still hormone sensitive. In the ORIOLE trial, Phuoc Tran did see an increased T-cell response after MDT after brief follow-up (see here), but I doubt it could be sustained without an immune checkpoint blocker.

    I would love to see your suggestion put to the test in a clinical trial. I was nonplussed that Karnes left M1a out of his MVA analysis. How did the journal reviewers let that pass?

  6. Tx for the reply, Allen.

    The whole concept of adding another treatment to make the tumor “hot” is intriguing and certainly warrants further study.

    Happy Hannuka — may all our lights shine bright for 2021.

  7. One other shortcoming of the study: the metastases were detected by C-11 choline scan, not the more sensitive PSMA PET scan. The ORIOLE study showed that irradiating all of the PSMA/PET positive lesions was far superior to only irradiating the conventionally detected lesions, so in a sense the Mayo study was only testing the effectiveness of the Oriole study’s non-PSMA control group.

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: