PSA After Prostatectomy: Tracking Changes Over Time

After a radical prostatectomy, PSA should fall to undetectable or near-undetectable levels because the organ that produces most of this protein has been removed. Tracking that decline and every subsequent measurement is one of the most reliable ways to gauge whether treatment succeeded, whether cancer has returned, and how aggressively any recurrence is behaving. But “undetectable” is not one fixed number, the timeline to get there varies more than many patients expect, and the speed at which PSA rises again matters at least as much as whether it rises at all.

How Quickly PSA Clears After Surgery

PSA does not vanish from the bloodstream overnight. The protein has a measurable half-life, and research shows that the clearance curve is not a simple straight line. One study found a two-phase pattern: an early rapid drop with a half-life of roughly an hour and a half, followed by a slower terminal phase where the half-life stretched to about two to three days for total PSA.¹ A separate study showed that roughly 60% of men followed a single smooth decline with a half-life around two and a half days, reaching undetectable levels within a month. The remaining 40% displayed a two-component curve where a second, slower phase could drag the half-life out to more than a week, and far fewer of those patients had cleared their PSA by day 28.²

The practical takeaway is that checking PSA too soon after surgery can give misleading results. A detectable reading two or three weeks out does not automatically mean cancer was left behind. Many clinicians wait six to eight weeks before drawing the first post-operative PSA, but even that standard window may be premature for some men.

Persistent PSA and the Risk of Jumping to Conclusions

A PSA that never drops below 0.1 ng/mL after surgery is generally labeled “persistent,” and it has traditionally been treated as a sign that curative surgery fell short.³ The concern is warranted: patients with persistent PSA show higher death rates compared to those who reach early undetectable levels.⁴ But a recent large study in JAMA Oncology found something unexpected. Among men whose pre-surgery PSA was very high (above 20 ng/mL), a detectable PSA at the typical six-to-eight-week check did not always carry the same grim prognosis. Many of these men had simply been tested before the protein had time to fully clear. When treatment was delayed and PSA was re-checked, a portion went on to reach undetectable levels on their own. The study’s authors argued that measuring too early, and then rushing into additional therapy based on that reading, risked overtreating patients who did not yet have evidence of true residual disease.⁵

This does not mean a persistent PSA should be ignored. It means the timing of the measurement matters. For men who had high PSA before surgery, a second or third draw over the following months may be needed to distinguish slow clearance from actual residual cancer.

What Counts as Biochemical Recurrence

Once PSA reaches undetectable levels, any confirmed rise is a red flag. But the exact threshold that defines “recurrence” is surprisingly inconsistent across the medical literature. The most commonly cited cutoff is 0.2 ng/mL, confirmed on a second measurement. Other researchers have used 0.4 ng/mL, or two consecutive rises of any amount, or various other criteria. A study examining 12 different published definitions found that, reassuringly, the choice of definition does not dramatically change the estimated importance of other prognostic factors like tumor grade or stage.⁶ Still, the lack of a single universal standard means you may see different numbers depending on which institution or study you are reading.

Ultrasensitive PSA Testing and Earlier Warning

Standard PSA assays have a detection floor around 0.1 ng/mL. Ultrasensitive assays can measure down to 0.01 ng/mL or lower, and they are increasingly used in post-prostatectomy monitoring because they can flag problems months before the traditional 0.2 ng/mL threshold is crossed.

One study found that an ultrasensitive reading of 0.03 ng/mL or higher on the first post-operative draw independently predicted later biochemical relapse, with roughly an eight-fold increase in risk. Using that lower threshold instead of the conventional 0.2 ng/mL provided a median lead time of about 18 months, meaning recurrence was identified a year and a half earlier.⁷ Another study found that a nadir as low as 0.01 ng/mL at the first monitoring visit could stratify patients into low- and high-risk groups for recurrence within three years, and that tracking ultrasensitive values over the first three to five years after surgery outperformed a single nadir reading in predicting later relapse.⁸

The flip side is anxiety. A detectable ultrasensitive PSA in the trace range (say, 0.01 to 0.02 ng/mL) may never progress. Some of those readings reflect non-cancerous sources of PSA or assay noise. Whether to use ultrasensitive testing and how to act on it remains a conversation between patient and clinician, not a blanket recommendation.

Where PSA Can Come From When the Prostate Is Gone

A question that surprises many patients: if the prostate has been removed entirely, why would any PSA be detectable at all? The short answer is that the prostate is not the only tissue capable of producing small amounts of PSA. Periurethral glands, tiny structures near the urethra, have been shown to stain positive for PSA in immunohistochemistry studies, and trace PSA has been found in the urine of prostatectomy patients.⁹ Retained benign prostate tissue left behind during surgery is another theoretical source, though research suggests this is unlikely to produce measurable levels.¹⁰

These non-cancerous sources typically produce only trace-level PSA. A steadily rising PSA, especially one that climbs past 0.1 or 0.2 ng/mL, is far more concerning for cancer recurrence than a stable blip at the very bottom of what an ultrasensitive assay can detect.

PSA Doubling Time and Velocity Tell a Story Speed Alone Does Not

When PSA does start to rise after prostatectomy, the raw number matters less than how quickly it is climbing. PSA doubling time, the number of months it takes for the value to double, is one of the strongest predictors of whether a biochemical recurrence will eventually become clinically dangerous.

A large study found that men with a doubling time under three months had more than five times the risk of developing metastases compared to those with a doubling time of ten to twelve months.¹¹ The effect on cancer-specific death is even more dramatic. One analysis reported that a doubling time under three months carried a hazard ratio of 25 for dying of prostate cancer, compared to a hazard ratio of just 1.35 for a high Gleason score alone.¹² At the other end of the spectrum, men whose PSA doubled slower than every 15 months were more likely to die of something other than prostate cancer altogether.¹³

PSA velocity, the rate of rise per unit of time, also helps distinguish local recurrence from distant spread. When combined with pathologic information like tumor grade and stage, velocity can guide whether treatment should target the prostate bed with radiation or address systemic disease with hormone therapy.¹⁴ Comparing different methods for estimating doubling time, researchers have found that all common calculation approaches discriminate prostate cancer death at a similar level, and that the absolute PSA level at the time of relapse performs at least as well as doubling time for predicting mortality.¹⁵

Surgical Pathology and What It Predicts About Future PSA Behavior

The pathology report from surgery provides important clues about how likely PSA is to rise later. Positive surgical margins, meaning cancer cells were found at the cut edge of the removed tissue, increase the risk of biochemical recurrence. But not all positive margins carry equal weight. Research has shown that the Gleason grade of the tumor at the margin matters more than the grade of the main tumor itself. When higher-grade cancer (Gleason pattern 4 or 5) was present at the margin, the risk of PSA recurrence roughly doubled compared to lower-grade cancer at the margin.¹⁶ In another study, recurrence rates climbed from about 36% with grade 6 at the margin to over 80% with grades 8 through 10.¹⁷

Other independent risk factors for early PSA recurrence after robotic prostatectomy include advanced pathologic stage (pT3 or higher) and high overall Gleason score (8 or above).¹⁸ Knowing these features helps clinicians decide whether to watch PSA closely or discuss early additional treatment.

When PSA Rises and Imaging Steps In

A rising PSA after prostatectomy tells you that something is producing that protein. It does not tell you where. That is where imaging comes in, and the landscape has changed substantially with the arrival of PSMA PET scans. These scans target a membrane protein found on prostate cancer cells and can spot recurrences at PSA levels far lower than traditional CT or bone scans can.

In one head-to-head comparison, PSMA PET/CT detected disease in about 64% of patients after prostatectomy, while conventional CT and bone scan combined picked up recurrence in roughly 45%. About a quarter of patients had recurrences visible only on PSMA PET.¹⁹ A more recent study using extended field-of-view PET scanners reported detection rates climbing from about 67% at PSA levels below 0.2 ng/mL up to 95% at PSA between 1.0 and 2.0 ng/mL.²⁰

At very low PSA levels (0.2 ng/mL or below), detection rates are considerably lower. A German multicenter analysis using conventional PSMA tracers reported an overall pooled detection rate of about 30% at those low levels, with slightly better pickup in patients who had higher initial Gleason scores.²¹ PSMA PET is now well established for evaluating biochemical recurrence and is increasingly guiding treatment decisions, including whether to direct salvage radiation to the prostate bed, to a specific lymph node, or to hold off entirely.²² In another comparison, PSMA PET identified disease in over half of patients at a median PSA of 0.23 ng/mL, while conventional imaging at a similar PSA level was positive in only about 10%.²³

Why Salvage Treatment Timing Hinges on PSA Level

For men whose PSA rises after prostatectomy, salvage radiation therapy to the prostate bed is the most common next step. And the evidence is clear: earlier is better, measured by the PSA at the time treatment starts. A large study found that men who received salvage radiation at a PSA above 0.25 ng/mL had a roughly 50% higher risk of death from any cause compared to those treated at 0.25 or below. That elevated risk persisted at cutoffs up to 0.50 ng/mL but was not seen below 0.25.²⁴

This is why PSA kinetics and ultrasensitive testing are not just academic exercises. Catching a rise at 0.03 or 0.05 ng/mL rather than waiting for it to pass 0.2 gives months of additional lead time to plan and begin treatment. For men with persistent PSA after surgery, salvage radiation combined with hormone therapy has shown benefit. Among those treated early with this combination, outcomes like progression-free survival and overall survival were comparable to men whose PSA had initially cleared and then risen, suggesting that prompt action can compensate for the initially worse starting point.²⁵ Among patients with persistent PSA who received salvage radiation plus hormone therapy, the PSA level at six to eight weeks after surgery was a strong predictor of progression. Those with levels above about 2 ng/mL at that time point fared significantly worse than those below it.²⁶

The Problem With Switching Labs

One underappreciated issue in PSA monitoring is that not all PSA tests are created equal. Different commercial assays can return different numbers from the same blood sample. A study comparing five different PSA assays found clinically meaningful variability between them, which becomes a real problem when serial measurements from different labs are compared to gauge a trend.²⁷ If your first post-operative PSA was drawn at one hospital using one assay, and a follow-up was done at an outside lab with a different assay, a small apparent rise might be nothing more than measurement variation. The simplest way to avoid this is to have all your PSA draws done at the same lab using the same testing platform.

How Often to Check and Whether Less Could Be Enough

Guidelines generally recommend PSA testing every three to six months in the first two years after surgery, then every six to twelve months after that. In practice, adherence varies. A population-based study found that while 96% of men had at least one PSA test in the first year after prostatectomy, about a quarter went at least one full year without a test during the first five years. By year five, roughly 80% were still getting at least annual testing, but only about 55% were hitting the more frequent every-six-months schedule.

For men at low risk of recurrence, specifically those whose ultrasensitive PSA stays very low, there is growing interest in stretching out the testing schedule. The rationale is straightforward: if your ultrasensitive PSA is 0.01 ng/mL or less and stays stable, the odds of clinically significant recurrence in the near term are low. Researchers have suggested that these patients could safely move to testing every two to three years, which would free up clinic resources without compromising outcomes.²⁸ For everyone else, sticking to the guideline-recommended intervals remains important, especially during the first few years when most recurrences surface.

Machine Learning and Future Prediction Models

Predicting which men will experience biochemical recurrence after prostatectomy has traditionally relied on a handful of pathologic variables: Gleason grade, stage, margin status, and pre-operative PSA. These factors are useful but imperfect. Machine learning models are being tested as a way to improve accuracy. In one study, gradient-boosted tree models and neural networks both outperformed traditional statistical models in predicting recurrence after robot-assisted surgery.²⁹ These tools are still in the research phase and are not yet part of routine clinical practice, but they point toward a future where PSA trajectories, pathology data, and imaging results could be integrated into personalized risk scores that adapt over time as new data points come in.