Which PSA Changes Matter and Which Ones Don't - 071
A Guide to PSA Doubling Time, Biochemical Recurrence, and When to Act
TL;DR: PSA doubling time is very important for biochemical recurrence after surgery or radiation and in nonmetastatic castration-resistant prostate cancer, but only if calculated correctly with 3+ tests over 3+ months. Fast doubling (<3 months) = 27× higher death risk. Always verify that your doctor used a validated calculator.
Changes in prostate-specific antigen (PSA) help doctors track prostate cancer. This post reviews how PSA levels change across disease stages and what these changes may indicate.
We examine PSA doubling time (PSADT), which estimates how long it takes for PSA to double and is usually, but not always, a reflection of how rapidly the cancer is growing.
We also cover common PSA changes, such as the temporary PSA bounce that occurs in some men after radiation and the rise that occurs when starting certain therapies.
I also explain how PSA behaves after stopping hormone therapy and why some aggressive prostate cancers grow without making any PSA at all.
Common abbreviations
PSA (Prostate-Specific Antigen): A protein made by cells in the prostate gland.
PSADT (Prostate-specific doubling time): The amount of time it takes for a PSA level to double.
AUA (American Urological Association): A leading medical organization that establishes standards for urologic healthcare.
NCCN (National Comprehensive Cancer Network): A non-profit alliance of cancer centers that develops guidelines for treatment.
MRI (Magnetic Resonance Imaging): A scan using magnets and radio waves to create detailed images of organs like the prostate.
ADT (Androgen Deprivation therapy): Treatment that suppresses male hormones to slow the growth of prostate cancer.
Important points for PSADT
Prognostic Tool: PSA doubling time (PSADT) is one of the most reliable methods for estimating risk when PSA begins to rise again. It’s especially useful in two situations where imaging still looks normal:
Biochemical recurrence after surgery or radiation, when the cancer has not spread and is still hormone-sensitive, and
Nonmetastatic castration-resistant prostate cancer (nmCRPC), when the PSA is rising despite hormone therapy, but scans remain negative.
In both situations, a fast PSADT is a standard reason to intensify treatment, because it identifies men most likely to develop visible metastases sooner.
Predicting Future Risks: After surgery or radiation therapy, a confirmed PSA rise without evidence of cancer on imaging is termed biochemical recurrence (BCR).
In BCR, PSA doubling time is one of the strongest predictors of whether the cancer will spread and become life-threatening.
A landmark study looked at men with BCR after radical prostatectomy. It found that compared with men whose PSADT was 15 months or longer, men with a PSADT under 3 months had about a 27-times higher risk of dying from prostate cancer.
Standards for measuring PSA changes
PSA doubling time is the number of months it takes for the PSA level to double. This tool assumes that the cancer grows at a steady rate over time. To get an accurate number, you must have at least three PSA results over at least three months. These tests should be at least four weeks apart to avoid being fooled by random fluctuations.
If you use only two results or examine a very short time interval, the final number will be unstable and will not reflect how rapidly the cancer is moving.
Several factors can make these numbers inaccurate. Testosterone levels should be stable because they directly affect PSA production. Medications such as 5-alpha reductase inhibitors (finasteride and dutasteride) or herbs (saw palmetto) must remain at the same dose throughout the measurement period.
Pearl: If you have been on a 5-alpha reductase inhibitor for at least six months, multiply the PSA by 2 to get the correct PSA level.
Patients should also use the same laboratory for all PSA tests to avoid errors arising from different testing methods.
When PSA does not show the whole picture
PSA is a very helpful tool, but it does not always show when the cancer is growing. In certain conditions, the cancer can spread even if the PSA remains low or does not change.
Some men have a very aggressive type of cancer from the onset that never makes much PSA. These tumors lack the features that allow regular prostate cells to produce PSA.
Some studies show that up to 15% of men with a PSA < 4.0 ng/mL actually have cancer when they get a biopsy. Relying only on PSA changes in these men can lead to a dangerous delay in finding the disease.
Pearl: The key is that PSA doesn’t work in isolation. A ‘normal’ PSA of 3.0 ng/mL means something very different for a 45-year-old man with a suspicious lump on exam than for a 75-year-old with a smooth prostate and stable PSA for 10 years. These factors are why doctors use risk calculators that combine multiple factors before recommending a biopsy.
In metastatic castrate-resistant prostate cancer (mCRPC) after years on hormone therapy, treatments can sometimes cause the cancer to change into a different type of cell called treatment-emergent neuroendocrine prostate cancer (t-NEPC). These prostate cancer cells usually do not produce PSA.
Signs of this change include new pain or new growth on imaging, while the PSA remains unchanged. Doctors must watch for these patterns in late-stage disease to avoid missing this type of cancer.
Red flags for t-NEPC include rising LDH with stable PSA and elevated chromogranin A or synaptophysin levels on biopsy samples.
Active surveillance
Men with low-risk cancer often choose to monitor the disease instead of having immediate treatment. In the past, a rapid doubling time was a common reason to stop active surveillance (AS) and start treatment.
However, studies show that PSA changes do not always correlate with microscopic findings. Many men have PSA changes because of an enlarged prostate or inflammation.
Modern guidelines now favor the use of MRIs and genetic tests to help decide if a cancer is high-risk and needs treatment or not. A rapid doubling time is now used as a prompt to get a new scan or biopsy, but not a stand-alone reason to initiate treatment.
NCCN recommends asking about family history and considering germline testing for hereditary mutations, such as BRCA, for specific patients at the time of initial diagnosis. In addition, tumor-based genetic tests, such as Oncotype DX Genomic Prostate Score, Prolaris, or Decipher, may help determine whether AS is appropriate for select patients.
Biochemical recurrence after definitive treatment
The definition of when cancer recurs after surgery varies among medical groups. The AUA defines biochemical recurrence as a PSA rise to ≥0.2 ng/mL with a confirmatory value >0.2 ng/mL.
The NCCN guidelines call it PSA Recurrence and define it as an undetectable PSA after surgery, followed by a detectable PSA that increases on two or more determinations or reaches a PSA > 0.1 ng/mL.
Both the AUA and NCCN agree that the optimal time to initiate radiation for biochemical recurrence is when the PSA is between 0.1 and 0.2 ng/mL. And that the most crucial deadline for initiating radiation for biochemical recurrence after surgery is before the PSA moves above 0.5 ng/mL.
Delaying treatment until PSA > 0.5 ng/mL is associated with poorer outcomes; however, the NCCN notes that patients with very slow PSA doubling times (> 24 months) may consider delaying treatment.
Biochemical failure, signalling the recurrence of cancer after radiation therapy, is defined by the Phoenix criteria: a PSA increase of 2 ng/mL or more above the nadir (the lowest PSA level achieved).
PSADT as a gatekeeper
Before 2018, men with high-risk (PSADT of 3 months or less) nonmetastatic castration-resistant prostate cancer had no therapy proven in phase 3 trials to delay metastasis or extend survival. Standard care was continued androgen deprivation therapy with observation, and older add-on approaches showed at best modest activity without a clear survival benefit.
Then three landmark trials arrived in rapid succession, and all three used a PSADT of 10 months or less as the entry requirement. This cutoff wasn’t arbitrary. Shorter doubling times are strongly associated with increased risk of metastasis and death, and the median baseline PSADT across all three trials was under 5 months.
Across all three studies, the primary question was whether intensifying androgen receptor blockade could delay the transition from invisible to radiographically visible disease.
The answer was yes. SPARTAN (apalutamide), PROSPER (enzalutamide), and ARAMIS (darolutamide) each demonstrated substantial reductions in metastasis and death, extending metastasis-free survival by approximately 2 years compared with ADT alone, and all three ultimately showed improvements in overall survival as well.
Pearl: If you have nmCRPC and your PSA is rising despite hormone therapy and your PSADT is 10 months or less, you are exactly the patient these trials were designed for, and three FDA-approved options now exist to discuss with your doctor.
PSA bounce
PSA changes after radiation to the prostate can be due to a phenomenon known as PSA bounce. PSA bounce is a mild, temporary rise in PSA that occurs in 30-33% of patients around 17-24 months after radiation, then drops back down without additional treatment.
Treatment and recovery transitions
The timing of tests is critical when a patient starts or stops a drug. Sometimes, starting a new treatment can temporarily increase PSA levels.
PSA changes can be hard to read when a patient stops ADT. Often, PSA levels increase as the ADT wears off and testosterone levels recover. However, testosterone production doesn’t always return to baseline, and providers should not use PSA values during this rebound period to calculate doubling time.
Taxane chemotherapy (docetaxel, cabazitaxel) and abiraterone can cause an early but temporary rise in PSA. Thus, PSA changes during initiation of these therapies can be misleading, and a PSA rise alone should not automatically trigger treatment discontinuation, provided the patient is clinically stable, feels better, and appears better on imaging.
The problem with guessing PSA doubling times
The PSA doubling time is effective only when used correctly. A study presented at a major 2025 medical meeting showed that many doctors do not use formal calculators to determine PSADT.
Researchers found that doctors had not documented doubling time in the medical record for 63% of men with high-risk nonmetastatic castrate sensitive prostate cancer and biochemical recurrence. Those who did record a number in the chart often failed to use validated digital tools for verification.
Among men whose physicians reported a doubling time in the medical record, the reported values were longer than the retrospectively calculated values in 88% of cases. This overestimation suggests that physicians may underestimate how aggressive a tumor actually is by relying on gut feeling instead of a calculator.
The doctors who guessed the PSADT instead of using a calculator thought the cancer was moving more slowly than it really was. This finding represents a serious problem that affects the speed with which patients receive life-saving care.
Men with properly calculated doubling times in their charts were three times more likely to receive appropriate treatment in a timely manner.
Pearl: Ask your doctor if they used a formal calculator to determine your PSADT.
Until the next newsletter, I wish you good health and much love.
Keith
This is an excellent “PSA literacy” post, especially your emphasis that PSA kinetics matter more than any single PSA value once you’re in the recurrence / post-treatment setting.
A few takeaways I hope every clinician and patient internalizes:
1. PSA doubling time (PSADT) is only meaningful if calculated correctly—≥3 PSA values over ≥3 months, spaced far enough apart to avoid noise masquerading as biology. Otherwise we’re just fitting a line to randomness. 
2. The risk stratification point is sobering but actionable: very fast PSADT (<3 months) identifies a truly high-risk biology (your “27× higher death risk” comparison really lands). 
3. Your practical thresholds are exactly what patients need: don’t “watch and wait” into a worse window when salvage therapy is on the table (and conversely, don’t overreact to a benign PSA bounce after radiation). 
4. And thank you for noting the nuance many miss: some aggressive cancers decouple from PSA, so “low PSA” is not always reassurance when the clinical picture or imaging suggests otherwise. 
The overall message is beautifully clinician-grade and patient-usable: track trajectories, validate the math, and match the urgency to the underlying kinetics, not to anxiety, anecdotes, or a single lab draw.
I am so DEEPLY grateful for your explanations. You have that rare gift of being able to relay deeply technical information and still make it understandable for the non-professionsl. 5 years after radical prostatectomy an oncologist at a major cancer center here told me to start 2 years of ADT (relugolix) and 6 months of radiation based on a PSA rise from .06 to .14 over 12 months and a "suspicious cloudy area" on an MRI. He didn't tell me the side effects of ADT other than "it may cause some hot flashes". A second opinion consult with an oncologist and radiation oncologist at a different major cancer hospital explained the full range of the devestating side effects that 2 years of ADT would have on a 78 year old man and the flimsy evidence on a new MRI. Based on their recommendations, I didn't start ADT and my PSA has fallen to .10 over the past 12 months, so I have been able to lead a normal life to get my life organized for when I actually do have to start ADT. And I was elated to learn from your last post about DECIPHER testing. I took this info to my urologist/prostate surgeon and was disappointed to learn that after taking 3 sets of transrectal biopsies over 2 years (total of 36 cores) and radical prostatectomy, no samples were ever submitted for DECIPHER testing. He was slightly dismissive of the usefulness of this test but said my tissue samples likely still existed and he would order the testing. You have taught a lot of us to keep our eyes wide open, how to ask informed questions, and trusted that non-professionsl could comprehend granular information.
From a carpenter in Evanston, Illinois.