Low detectable PSA after prostatectomy – watch or treat?

In a previous article, we looked at evidence that a low detectable level of PSA predicts eventual biochemical recurrence (a confirmed PSA greater than 0.2 ng/ml) when there is aggressive pathology. But what is one to do when the pathology report is not necessarily poor (that is, the cancer may be fully contained within the prostate, and all surgical margins may be negative), yet the PSA is detectable and possibly rising?

Because several randomized clinical trials have demonstrated an advantage to earlier treatment over waiting, the National Cancer Center Network (NCCN), which comprises many to the top US cancer centers, uses a lower threshold for defining biochemical recurrence:

  • PSA detectable (≥ 0.03 ng/ml) after prostatectomy, or
  • PSA undetectable after prostatectomy that is subsequently detectable on at least two PSA tests

The NCCN definition may lead to over-treatment of patients in whom the small amounts of PSA may be attributable to benign tissue left behind, extraprostatic sources, or indolent cancer that may never progress in the patient’s lifetime. On the other hand, waiting for the American Urological Association (AUA) definition of a confirmed PSA greater than 0.2 ng/ml may allow the cancer time to progress beyond the local area.

Koulikov et al. of Roswell Park Cancer Institute addressed this problem in a retrospective study published in the Journal of Urology. They wanted to determine whether the pattern of low detectable PSA during the first 3 years after surgery could be used to predict eventual biochemical recurrence (a confirmed PSA ≥ 0.2 ng/ml). Their institutional database analysis was based on 556 prostatectomy patients treated between 1993 and 2008 whom they assigned to three groups defined as:

  • Undetectable PSA” (419 patients)
    • 0.03 ng/ml or less
  • Low detectable, stable PSA” (93 patients)
    • PSA greater than 0.03 and less than 0.2 ng/ml
    • No two subsequent increases in PSA, and/or
    • PSA velocity less than 0.05 ng/ml/yr
  • Low detectable, unstable PSA” (54 patients)
    • PSA greater than 0.03 and less than 0.2 ng/ml
    • Two subsequent increases in PSA, and/or
    • PSA velocity of 0.05 ng/ml/yr or greater

The primary endpoints they looked for were either biochemical recurrence (a confirmed PSA ≥ 0.2 ng/ml) or salvage radiation therapy beyond 3 years of follow up. They could not draw any meaningful conclusions about survival because of the relatively short follow up.

The 7-year recurrence-free survival rates for the three groups were found to be:

  • 95 percent in the “undetectable PSA” group
  • 94 percent in the “low detectable, stable PSA” group
  • 37 percent in the “low detectable, unstable PSA” group

The post-surgical pathological findings of the “undetectable PSA” and the “low detectable, stable PSA” groups were nearly identical, while there were significant differences compared to the “low detectable, unstable PSA” group. Intermediate- and high-risk patients were more often found in the “low detectable, unstable PSA” group compared to the other groups.


The presence of a “low detectable, unstable PSA” was a significant predictor of biochemical recurrence, along with pathological stage, Gleason score, and positive surgical margins. It would be useful to know if those patients progressed to biochemical recurrence even if they did not have aggressive pathological characteristics; however, with only 54 patients, it would be impossible to draw reliable conclusions.

If these findings are confirmed in randomized clinical trials, post-prostatectomy patients with “undetectable PSA”, or “low detectable and stable PSA”, could be safely watched.

There is an open controversy as to whether salvage radiation therapy, even if given after biochemical recurrence (a confirmed PSA ≥ 0.2 ng/ml), translates to a survival benefit. Fewer than a third of patients with a post-prostatectomy biochemical recurrence experienced systemic progression, and it takes a median of 8 years for distant metastatic progression, and 13 years for mortality to occur, according to a Johns Hopkins study (by Pound et al.). This study may help inform patient and doctor discussion and choices.

Editorial note: This commentary was written for The “”New” Prostate Cancer InfoLink by Allen Edel. We thank Dr. James Mohler for providing access to the full text of the study by Koulikov et al.

29 Responses

  1. I am one who has chosen to wait and see, I have had three PSA tests since my RP on 1/10/14 and a fourth due this month. I have been undetectable so far, even though I had one positive margin. I had two opinions from some great doctors and both said it was a 50-50 decision to go ahead with radiation. That was not enough for me to go ahead; they both have supported my decision to monitor closely.

  2. My previous article showed that among men with a positive margin or stage pT3 none of them went on to biochemical recurrence as long as their ultrasensitive PSA remained undetectable or < 0.03 ng/ml. That should give you and your doctors a lot more confidence in your decision to watch.

  3. I’ve been > 0.04 to > 0.05 for 8 years now. My first PSA was > 0.04 ng/ml. According to the article I should have had adjuvant radiation 8 years ago if I am reading this correctly?

  4. Dear Ron:

    The distinctions between a “low, stable, and undetectable” PSA level and a “low, unstable, and undetectable” PSA level can be very subtle. These subtleties can profoundly impact the appropriate management of individual patients. In my entirely personal opinion, you appear to have a low, stable, and undetectable PSA level that has never required any form of adjuvant or salvage therapy. The findings in the study discussed above are findings from one study. They suggest implications, but they don’t tell us how to treat individual patients in any absolute sense.

  5. Excellent review of an intriguing paper. Any study helping to counter the “aRT for everyone” trend is welcome.

    I might point out that at least one of the findings is unsurprising. If PSA starts very low and does not rise significantly, the patient will never reach BCR, or major symptoms of prostate cancer. He should not need advice or guidelines to decide to spare himself further treatment. If PSA is rising, the decision turns into a game of high-stakes chicken. The statement “… waiting for the American Urological Association (AUA) definition of a confirmed PSA greater than 0.2 ng/ml may allow the cancer time to progress beyond the local area” is equally true for other values of PSA and does not enlighten the decision.

    As you noted, the timing and efficacy of sRT is controversial. Studies have used varying cutoff levels for “early” sRT from 0.03 to 1.0 ng/ml, making them difficult to compare. One might ask whether it is useful to worry about having an early warning of BCR, when there is no proof that early sRT is better than later sRT, or whether sRT at any level translates to a survival advantage, or even whether BCR is a surrogate for later undesirable events. Once again we must await further studies.

  6. My husband had a prostatectomy on May 12, 2014: margins negative, Gleason 7, past 5 months and five PSAs ranged between 4s and 5s.

    On October 10, 2014, he got a 6-month Lupron injection. Now, as of January 8, 2015, his PSA is 0.02 ng/ml. His cancer was all microscopic and was also in one seminal vesicle. Should he do the radiation? Our doctor says if the Lupron is working then no. But we hear that radiation would be the only cure, but they will radiate blindly because they can’t see it anywhere on any scans. Please help.

  7. Ron,

    I’m not sure you are reading it correctly. The study found that progression occurred only when the PSA was rising or “unstable.” Yours appears to be quite stable, which would put you in the second category. Almost all patients in that category (94%) enjoyed progression-free survival out to 7 years of tracking.

  8. Dear Karen:

    Your husband needs to consider asking about having some form of imaging scan that may be able to actually tell whether his cancer is in an area that can be radiated or not.

    You don’t tell us what his PSA was prior to surgery and you also don’t say whether he had any positive surgical lymph nodes (or whether the lymph nodes were sampled but were all negative). I also am not clear what you mean when you say that his cancer was all “microscopic” — microscopic where?

    The types of imaging test that might be able to help your husband understand where his cancer remains are things like the choline-C11 PET/CT scan done at the Mayo Clinic in Rochester, MN, and some types of sophisticated MRI scan available at major specialty medical centers.

  9. Allen,

    Nobody thought I would be undetectable for this long with the palpable nodule that was felt along with a post-op Gleason 7 with 40% gland involvement. Guess that is why I always still sweat and keep close track of my PSA for last 8 years

  10. Karen,

    I’m not clear from your post exactly what you’re saying about his PSA after surgery. Are you saying his PSA was between 4 and 5 on his post-surgery PSA tests? That would explain why his doctor gave him a Lupron shot, which brought his PSA down. When you say “it was micrometastatic,” I think you mean that no metastases were visible on a bone scan? His pathological seminal vesicle invasion is a serious adverse finding.

    Both the high PSA post-surgery and the seminal vesicle invasion increase the risk that the cancer is systemic and can’t be cured with radiation to the prostate bed. Nevertheless, you may decide that you want to try it anyway. That’s a conversation you should be having with a radiation oncologist.

    There aren’t any known forms of imaging that are likely to be able to find metastases while his PSA is that low from the Lupron.

  11. Once again, terrific write-up, Allen. I think the InfoLink has made a practical decision to include your insight. Your ability to break down the data is amazing to me!

  12. Aw shucks, Tony.

  13. I’m endorsing Tony’s observation, Allen … very helpful to me and to men I am mentoring … Tx!

    I had always thought around 30% of men experienced BCR (recurrence) post-RP although the Johns Hopkins study suggests around 15%. What’s your understanding?

    I believe that Peter Carroll’s lower PSA threshold is a PSA of 0.075 rather than 0.03, although he may have lowered that since I last spoke to him.

    Tx again,


  14. Thanks. That’s very kind of you to say, Rick.

    I think it depends on the time frame. Johns Hopkins is telling us the BCR is 15% at 5 years, but elsewhere they report progression-free survival is 68% (32% recurrence) at 25 years. Also, we expect Johns Hopkins to do better RPs (they have a low positive surgical margin rate) and to do a better job at choosing appropriate patients compared to the average institution. So the 30% figure that you and I have seen is probably a population-wide figure with no specific time cut-off.

  15. Have had the surgery in 2013 and then had to have radiation in July of 2015. PSA has been good. Four months ago was 0.03 and today 0.05 is there any need to be concerned at this time?

  16. Charles:

    That’s a question you need to ask your doctors, but they would probably tell you that it rather depends on what your next PSA level is. It might stay the same. It might go down again. Or it might continue to rise. The last would suggest a possible need to think about what comes next.

  17. I see here that the threshold number for detection is 0.03 ng/ml, and that is what I was told by a first screening doctor. My later doctor says the number is 0.06 and below is undetectable.

    What is the true number to use? And is there really that much difference between the two numbers if one’s PSA stays at, say, 0.06 ng/ml after two consecutive PSA testings over 6 months?

    My confusion is that different doctors use different numbers as what constitutes undetectable.

  18. Michael:

    The minimal number that can be assessed accurately as a “standard” PSA level from a standard blood draw depends on the exact test being used and the rigor with which the laboratory uses that test. Some ultrasensitive tests can read PSA levels down to 0.003 ng/ml or lower.

    Having said that, the important thing is to get all one’s PSA tests done by the same laboratory if at all possible. Then one is going to be able to accurately compare “apples to apples”.

    If one’s PSA level was being assessed using the same three consecutive blood samples at three different laboratories, one might get three different sets of results, like these:

    — Lab A might give you readouts of 0.030, 0.035, and 0.030 in January, April, and July.
    — Lab B might give you readouts of 0.03, 0.04, and 0.03 in January, April, and July
    — Lab C might give you readouts of <0.1, < 0.1, and < 0.1 in January, April, and July

    All three sets of results actually say the same thing … that your PSA is low and stable.

    It is also worth knowing that there is really no such thing as a truly “undetectable” PSA level. There are PSA tests that can be used in research laboratories to test for PSA levels down into the femtogram per liter range (as opposed to the normal nanogram per liter range). To understand what this implies, you need to know that 1 fg/ml = 1,000,000 ng/ml (i.e., these highly specialized tests are a million times more sensitive than the standard tests used to measure PSA levels). However, we don’t use these types of research laboratory tests in standard clinical practice. It would cost far too much. So, …

    Get your blood samples all sent to the same laboratory for PSA testing so that everyone is comparing apples to apples. If they keep coming back at about the same level and that level is below 0.1 ng/ml or better still at or below 0.03 ng/ml, then you are in good shape.

  19. Michael,

    Let me call your attention to this more recent article on a similar subject.

    As you can see, RADICALS, the randomized clinical trial, found that most patients could comfortably wait until their PSA had risen to 0.1 or three consecutive rises. You may use an ultrasensitive PSA test to count the three consecutive rises (every 3 months). You may want to start salvage radiation earlier if you have significant risk factors like a high post-op Gleason score; significant positive margins with high Gleason score at the margin; persistent PSA; seminal vesicle invasion; positive lymph nodes; or rare histology (e.g., ductal, IDC-P, neuroendocrine, etc.).

  20. Hi Allen. Actually the link was there but there was a “glitch’ in the HTML for some unknown reason. Sorry about that. It’s fixed.

  21. Hi Allen,

    I had my prostatectomy in February 2020. My PSA was 5.3 before the operation. The pathology after the operation showed the following:

    — Gleason 9
    — Margins were zero
    — No lymph nodes were removed

    My PSA 1 month after the operation was 0.04 and kept moving for the last 6 months between 0.04 to 0.048. This month, September, which is the 7th month from the operation date, my PSA went up to 0.058. Do you think this is alarming or too early to reach any conclusions ?

  22. I think it’s too early.

    Since I wrote this more than 5 years ago, we’ve gotten some results from the RADICALS-RT trial. They showed that waiting for three consecutive PSA rises or waiting for PSA to reach 0.1 ng/ml gave equivalent results to immediate treatment. So you have nothing to lose by waiting for a more definitive indicator.

  23. Hi Allen.

    i had a radical prostatectomy almost 3 years ago. My first PSA result (3 months post-operation) was 0.01 ng/ml, then every 2 months for the first 2 years was rising by approx. 0.011 and reached 0.16 ng/ml by the end of 2019. January 2020 i had a PSMA scan that showed no clinical evidence of pc progression.

    Since, my PSA dropped to 0.15 ng/ml and I have had three subsequent tests (over the last 10 months) showing a stable PSA value of 0.15 ng/ml, which category do I fall into?

    PS: Pre-surgical data were as follows: PSA, 6 ng/ml; Gleason score 3 + 3 = 6. Post-surgical data were: Gleason score 3 + 3 = 6; cancer in just one lobe; one unifocal positive surgical margin; no seminal vesicle involvement.

  24. Hi panos.

    It is likely that some GS 6 cells were left behind. They don’t directly metastasize, but may, slowly over time, progress to something else that does.


    I think that you are in a similar situation to men on active surveillance for low-risk prostate cancer. Like them, I think you ought to have periodic mpMRIs, and if there is any indication of progression in your prostate bed, maybe biopsy that area.

  25. Hi Allen and thanks for answering.

    If cancerous cells increase over time, wouldn’t the PSA reading increase too?

    How often should i have the mpMRIs considering PSA does not exceed 0.2 or remains stable?

  26. Yes. It is as if you are on active surveillance for low-risk prostate cancer now: PSAs twice a year to start. MpMRIs on no set schedule, but dependent upon the last one — If negative, you may be able to go 3 years before the next one, and if still negative, maybe 5 years after that (if PSA is low and stable). But if you see a PIRADS score ≥ 3, more frequent mpMRIs and possibly fusion biopsies may be prudent. I suggest you enroll in an active surveillance program rather than trying to do this on your own. I know that MSKCC has a well-researched schedule.

  27. Hi Allen, and thanks again for your time and prompt answer. I have one last question.

    Considering that I had a PSMA PET scan in February 2020 (i.e., 8 months ago), which is considered to be the most accurate test for detecting clinical evidence of prostate cancer, when should I go for the first mpMRI scan?

    Regarding MSKCC I have seen in the past that they had some nice prediction tools / prostate cancer nomograms (e.g., this one) based on their own patient database.

  28. Hi panos,

    As I said, “I suggest you enroll in an active surveillance program rather than trying to do this on your own. I know that MSKCC has a well-researched schedule.”

    It is likely that the mpMRI will find no significant PCa. It is only good at finding cancer that is higher grade or larger in size. That’s why MSKCC’s mpMRI/biopsy protocol makes sense (this has nothing to do with their nomograms you linked). PSMA PET scans only find cancer that expresses PSMA, and I would guess that yours does not, so it is useless for this purpose.

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: