PSA is back down

My two previous readings were <0.014 and, 6 months later, 0.037. So my uro thought we should test again in 3 months rather than 6. I just got the results: 0.020. So myLabcorp uPSA history looks like:

(3 m ) 0.010
(6 m ) 0.009
(9 m ) 0.007 (nadir)
(1 yr.) 0.018 (huh?)
(13 m) 0.013 (retest)
(15 m) 0.012
(18 m) 0.015
(21 m) 0.015
(2 yr. ) 0.016
(28 m) 0.015
(34 m) <0.014
(40 m) 0.037 (?)
(43 m) 0.020

I think the 0.020 is (perhaps) just low enough to call the pervious 0.037 an outlier. Looking back, I had a bump to 0.018 at 1 year post-op. So the 0.020 isn't that much out of line with the approx. 0.015 I've been running recently.

Research has shown that the ROCHE uPSA assay itself under perfect lab conditions can vary ±0.002 (2 standard deviations = 98% of results) when measuring PSAs around 0.010. Of course you and I don't have our tests run in perfect conditions, so let's say it's more like ±0.003 or ±0.004 in daily operation. Then we don't know what normal, physiological variation is when you PSA is actually stable and not rising. How much does a "stable" PSA vary hour to hour or day by day? Then you might add in a little sloppiness in sample prep, or equipment calibration, and it isn't hard to see how you can get an outlier reading when all the intrinsic and extrinsic error sources happen to all go in the same direction on any given day. And yet, if you look at my post-op PSA history, the results are actually in a fairly narrow range, with the 0.037 exception.

We decided (at my visit 2 days ago when I had my blood drawn) that we should do the next test at 3 months regardless of today's result, just to have more data points. If the next comes back below 0.020, I'll probably go back to 6-month testing. It really does take a number of data points to establish an upward trend.

I had another health scare last month involving surgery. I posted it in another HW Forum.

Djin

halbert said...
Djin,

What you see is exactly why you want to see successive rises several months apart before jumping off the additional treatment cliff. I've argued this point forever in here--and what you write is one of the first statement's I've seen about the limits of precision on the PSA test. I'd be the one guy who wishes that Labcorp added the precision limits on every test--something that would hopelessly confuse the majority.

Labcorp (and every other lab to my knowledge) doesn't publish the limits of precision. I posted at length on my research into the ROCH PSA assay that Labcorp uses in another thread here some time ago. My information there was based in part on a fascinating study that was published in the journal Practical Laboratory Medicine:

The significance of reporting to the thousandths place: Figuring out the laboratory limitations (2017, Full Text)

where an independent lab looked at the value of reporting to the thousandth place and chose the ROCH uPSA assay (which Labcorp uses) as their test assay. It's a fascinating read -- but keep in mind that the are looking at the assay in ideal test conditions, performing 20 runs on a sample with a target (i.e. known) PSA concentration of 0.01. (They also did repeat runs on a target PSA concentration of 0.2).

They also looked at actual patient data. My take-away conclusion is that when you round the Labcorp 3-decimal uPSA test to 2 decimal places, you can have very reasonable confidence in the 2nd data place. This is unlike a 2-decimal test, where uncertainty remains in the (rightmost) 2nd decimal.

Djin

Post Edited (DjinTonic) : 4/3/2021 7:09:42 PM (GMT-6)

halbert said...
Djin,

Thanks. When I have time I'll read it. One of the things that is also involved here is what is really a philosophical question: What does it MEAN? Does the number in the thousandth's place mean anything? And what about the next generation of instruments that will be reporting the ten-thousandth's place? Where do we say enough?

That level of precision is already here for uPSA, where research assays go down to fractions of picograms! What does this mean for us? Nothing, just as blood pressure reading of 120.02/82.17 would be possible, but needless precision, given how BP varies from second to second and clinical decisions are based on much less precise, but perfectly adequate, readings.

My rule of thumb is for guys to use a PSA that has one decimal place more than the one they want to follow. Basically, a 2 decimal test is fine for most post-treatment men, but some high-risk, post-RP men may want a 3-decimal test if they want a heads-up for a rising PSA.

A precise, accurate uPSA reading can be helpful for knowing one's post-op PSA nadir. The nadir can be used to statistcally predict the likelihood of future BCR. Basically, a PSA nadir below 0.03 is good; below 0.01 better. A PSA at 3 years post op of 0.03 or less is also good.

Djin

Post Edited (DjinTonic) : 4/2/2021 1:12:28 PM (GMT-6)

Djin, thats good news, and hopefully the next PSA in 3 months will confirm it.

And glad that the other condition, which i read about in the Allergy forum, turned out to be benign.
As the great Gilda Radner used to say.... "its always something"

2021 is looking up!

Djin,

Been away a bit but see you are monitoring psa closely and being prepared. Good to have trend lines for clear-headed decision making.

Also had sinus surgery on one side a decade ago. Have been doing the saline rinses and nasacort/fluticasone to keep things open. Seems to help.

You never know what's coming health-wise at our ages.

Sometimes I really feel like Mr. Cage in this scene from Raising Arizona...

https://vimeo.com/185715310

Djin Tonic,
That is great news, I am sure that trends will continue with little variance. Great word you used there-"outlier". sounds like the title for a movie...or what my wife sometimes calls me ...an Out and out liar! lol
Hopefully that sinus issue resolves, unclogged breathing is ...priceless.
Take care, thanks for all your posts, and as always, keep the faith!
Wings aka Dan in So Cal