Your PSA is Undetectable

Ed Weber, M.D., Editor
January- February, 2015

The Measurement:

It used to be simple. An undetectable PSA pf <0.1 ng/mL 4 - 6 weeks after surgery carried an encouraging bit of reassurance of a clean removal and a good prognosis.

But … no more!

The more sensitive the assay, the shorter the period of this reassurance. The values have steadily lowered from concern at levels >0.01 (beginning in the 1990s), then to >0.006 or even to >0.008, and currently to >0.003 (the AccuPSA). That represents 3 picograms/mL — an amazing 3 trillionth of a gram! Tests that report results in picogram range are termed ‘fourth generation’.

Next came a remarkable radioimmune assay with a detection cutoff of <1 pg/mL.

Even more notable is a “fifth-generation” test employing a digital immunoassay with single-molecule counting, said to have a limit of quantification (LoQ) of <0.05 pg/mL!

Studies of each of these tests come replete with data on the prediction of biochemical recurrence in relation to values higher or lower than the test’s lower limit of detection.

PSA ‘Progression’: What Is the Appropriate Threshold for Declaring ‘Progression’?

In the ‘old days’ PSA progression after surgery, i.e. biochemical failure, was stipulated when a PSA rose above the comfortable but arbitrarily chosen level of >0.2 ng/mL. This continues as the definition currently used in reporting most clinical trials. A threshold of >0.4 ng/mL had even been suggested at the Mayo Clinic because of studies that showed that 50% of men between 0.2 and 0.4 ng/mL did not progress. Dr. Tom Stamey in l996 published data based on persisting PSA in men thought to be cured of prostate cancer after surgery. Using a test sensitive to >0.002 ng/mL, he concluded that only PSA values above 0.07 ng/mL should be regarded as indicating residual prostate cancer.

However, now PSA progression may be declared at any level that exceeds the lower limit of detection in any of these ultra- and very ultrasensitive tests, since the application of the term ‘undetectable’ carries the reciprocal implication that PSA is ‘detectable’ above the lower test limit. What clinical interpretation and management decisions should result from test values above and below these lower limits of detection?

What About Assay Variability in Sequential Tests?

The utilization of ultrasensitive assays, such as the remarkable lower limit of detection of <0.003 ng/mL, raises the question of the utility and interassay reproducibility, i.e as in sequential testing. Drs. Wilson, Partin, et al. address this issue in their article, “Fifth-Generation Digital Immunoassay for Prostate-Specific Antigen by Single Molecule Array (SiMoA) Technology” (Clinical Chemistry. Oct 2011 – free full text online). When discussing tests reported in the pg/mL range, i.e., <0.003 ng/mL, they indicate that the functional day-to-day limit of quantification may vary as much as +/- 20%.

Their motivation for pursuing a very ultrasensitive assay was evidence that “suggests, however, that more sensitive determination of PSA status following RP could improve assessment of patient prognosis and response to treatment and better target secondary therapy for those who may benefit most.”

Studies Relating Ultrasensitive PSA Values to Prediction of Biochemical Recurrence (BCR):

Current studies have considerably lowered the bar for designating the nadir PSA level post-RP that should raise concern. In his article using the ‘fifth generation’ SiMoA test with a detection sensitivity in the 0.001 ng/mL range, Wilson seemed comfortable in regarding men whose post-RP PSA was below 0.01 ng/mL (correct as written) as likely cured, especially if the low value remains steady over time. This would represent a considerably lower bar than the 0.07 ng/mL suggested by Stamey.

The maintenance of a post RP PSA below 0.01ng/mL was also endorsed by Doherty et al. (British Journal of Cancer 2000) as a useful indicator of relapse free survival: “Only 2 patients with an undetectable [i.e. <0.01 ng/mL] prostate-specific antigen after radical prostatectomy had biochemical relapsed (3%), compared to 47 relapses out of 61 patients (75%) who did not reach this level.” Their 5 year data suggests that maintenance of a PSA <0.01 for 2 years predicts a high likelihood of long-term disease free survival.

Partin et al. (BJU Int June 2012) have taken the issue to a further lower level. In a study of 31 patients they reported that 100% of men with a post-RP mean nadir AccuPSA of <0.003 ng/mL were free of biochemical recurrence compared (BCR) compared to 62% of men with a PSA above that level. “Those men without evidence of BCR had a minimum of 5 years’ PSA follow-up.” The mean nadir for non-recurrence was 0.0023 ng/ml compared to 0.047 ng/mL for those with BCR. Their conclusion: “The AccuPSDA assay predicts 5-year BCR-free survival after RP.”

All articles ( see Wilson, Doherty, and Partin, above) agree that PSA doubling times in the ultrasensitive range are unreliable for supporting clinical management decisions. This was especially addressed by Carroll et al. (J Urol.2011 Oct): “Poor Agreement of Prostate Specific Antigen Doubling Time Calculated Using Ultrasensitive Versus Standard Prostate Specific Antigen Values” [i.e., PSA doubling times calculated in the range >0.2 ng/mL].

“The only safe generalization continues to be that the lower the nadir PSA, the more likely the patient will enjoy long-term BCR-free survival” ( Wilson, ibid).

What is the Clinical Utility of the Ultrasensitive PSA tests?

Clearly a post-RP PSA value above various ultra sensitive thresholds, i.e. ranging from >0.01 ng/mL to >.003 ng/mL have probability implications for biochemical recurrence in men with positive surgical margins and for the presence of residual cancer.

Furubayashi et al. (Oncol Lett. Nov 2014) addressed the implication of ultrasensitive PSA values as regards positive surgical margins.They reported a study of 239 men with one or multiple positive surgical margins. Of those whose PSA nadir was <0.008 ng/mL 93% of men with single or multiple positive margins remained free of recurrence while 46% with a higher value had a BCR. The median follow-up period was 53.8 months. The risk for recurrence was 11X for men with multiple positive margins whose PSA nadir was >0.008 ng/mL vs lower.

In Urology 1997 Oct Drs. Ellis, Lange et al., assayed ( with a sensitivity <0.008 ng/mL) sera from 29 men post cystoprostatectomy and compared the results with men post radical prostatectomy as controls. The surgical patients had been free of recurrent disease for 5 years. PSA levels for cystoprostatectomy patients were <0.008 ng/mL in 86.2% and <0.008 ng/mL in 96% of the ‘cured’ prostatectomy patients. These data supported the findings of Furubayashi using the same low PSA value.

Limited Practical Applications Currently Available.

The available clinical applications based on these very low ultrasensitive assays currently do not match this exquisite assay technology. Salvage radiotherapy would likely be chosen as the appropriate treatment when the prediction of recurrence is substantial. However, in their 2007 JCO study, “Predicting the Outcome of Salvage Radiation Therapy for Recurrent Prostate Cancer After Radical Prostatectomy,” Stephenson et al. supported initiating radiation treatment at PSA levels of 0.5 ng/mL or lower, a value that is 2 logs higher than ultrasensitive levels. Current practice of initiating salvage radiation therapy follows that guideline today.

It is possible to speculate that immunotherapy, which is considered best applied when the tumor burden is lowest, may become an effective modality to be prescribed for men in whom BCR is predicted based on ultrasensitive assays. But the state of the art for this modality is not nearly ready to address this task.

BOTTOM LINE: It is becoming standard to measure postoperative PSA values with the almost unbelievable sensitivity of several picograms. Clearly values in this range can provide predictive information and can be satisfying (or not) for the surgeon. PSA values above or below a stipulated value can be a source of reassurance — or anxiety, for men, especially when a value might be edging upward in picogram increments. However, the state of the oncological art is not ready to effectively accommodate these very low values. Until then, a rise in PSA above 0.01 ng/mL is the value to appropriately raise concern.