What does it mean?
PSA is undetectable when it is too low to be detected by the assay system being used.
The Prostate Cancer Research Institute (PCRI) defines it as "a PSA of <0.05 using a hypersensitive assay such as DPC Immulite 3rd generation PSA or Tosoh assay."
Johns Hopkins defines it as less than 0.2 ng/ml.
Walsh says that with
levels as high as 1.0 to 1.5 ng/ml, this means that 10-15 grams or more of viable prostate are present (the PSA level in the serum is approximately 10% of the weight of the prostate).
Subject: Re: [PP] What is undetectable PSA?
> Date: Wed, 6 Mar 2002 17:57:14 -0500
> From: Don and Winnie Janzen
> Subject: What is undetectable PSA?
> Could someone one please tell me what is considered "undetectable PSA =
> reading". My husband Don had a RP last June and had two checkups since =
> where his PSA was "undetectable". He just had another PSA reading done =
> at another hospital and it came back as 0.1. We want to know if that is =
> considered "undetectable"??? Or, could the reading be different at a =
> diff. hospital?
> We are worried.
As I see it, "undetectable" as related to PSA is used to convey two
somewhat different meanings.
>From a strict technical point of view, an "undetectable" reading is one
where the equipment failed to detect the marker sought, PSA in this
case. These "undetectable" readings are usually reported as "less than X ng/ml",
where X is the "detection limit" of the lab, assay, or machine used. As
I recall, X = 0.2 ng/ml for Hybritech (one of the original assays, still
in use), 0.03 ng/ml for Tosoh (one of the first high sensitivity
assays), and 0.002 ng/ml for the DPC IMMULITE 2000 (one of the
ultrasensitive "third generation" assays). Therefore, for example, a
true PSA level of 0.1 ng/ml would be reported as "undetectable" (less than 0.2
ng/ml) by a lab using the Hybritech but would be reported as
"detectable" by a lab using the Tosoh or the DPC 2000. As you know, to
properly assess the PSA trend all readings should be obtained from the
same lab using the same assay or machine type.
A different perspective arises when "undetectable PSA" is loosely used
to describe the expected outcome from a definitive local procedure (i.e.
one which is expected to totally eradicate the tumor by completely
removing or destroying it). For example, "a successful RP should result
in an undetectable PSA". Here, there are other factors that must be
taken into account. First, even if the surgery removed the entire gland
and there was no prostate cancer elsewhere (i.e. there had been no
spread of the cancer beyond the prostate prior to the surgery), there
could still be minute amounts of circulating PSA produced by other
tissues which would be detectable by the ultrasensitive assays. This is
why, according to Dr. Stamey from Stanford (one of the preeminent
prostate cancer surgeons) when using an ultrasensitive PSA assay (such
as the DPC 2000) one need not start to worry about a possible
"recurrence" unless the PSA is detected above 0.05 ng/ml (notice that
the Hybritech assay would still report this as "undetectable", or <0.2
In addition, as I understand it, the PSA could be higher than 0.05 ng/ml
and still not represent a "recurrence". For example, it appears that it
is not uncommon for the surgeon to inadvertently cut through the edges
of the gland as it is removed (i.e. some of the gland might have been
left behind). When the tumor extends to that area the result are
"positive margins" and a higher probability of recurrence. But when the
portion of the prostate left behind is benign, the PSA might well be
above 0.05 ng/ml because the prostate tissue left behind might continue
to produce PSA and/or even regenerate, even though there is no tumor
left. As I see it, the key difference is that if tumor was left behind
the PSA will not only be "detectable" but it will continue to increase
in a roughly exponential fashion (unless the residual tumor withers by
itself due to the lack of blood supply or via the activities of the
immune system), whereas if only benign tissue is left behind the PSA
will be detectable but will either remain mostly stable or, at most,
will increase slowly in a more linear fashion.
When the definitive treatment is radiation a different situation exists.
As I understand it, therapeutic radiation relies in the expectation
that, although both the cancerous and the benign cells might sustain an
equivalent amount of damage to their DNA, the cancerous cells will be
more unlikely to be able to repair themselves (because of their already
defective DNA) and, thus, the radiation will tend to selectively kill
more cancerous cells than benign cells. The aim of therapeutic RT is to
apply a dose that will kill all of the cancerous cells but not
necessarily all of the benign cells. This, IMO, explains why a
"successful" RT might result in a stable PSA well above the PSA level
expected from a "successful" RP. It also means, IMO, that it is not
proper to compare the "success rates" between RP and RT by merely
comparing their PSA outcomes if these outcomes are rigidly expressed as
whether they are above or below any given threshold. As I say, as long
as the PSA remains stable there is reason to hope that cancer is not
present. Of course, given how long it takes some times for a residual
tumor to manifest itself as a rising PSA, one can't be sure until many
years have passed.
I am sure there is more to it than what I posted above but I hope this
might help someone. Any constructive comments are welcome.
I am not a medical doctor. The foregoing is not medical advice. It is
only my opinion. You should seek other opinions from qualified licensed
The material included here is intended for educational purposes only,
and must not be considered a substitute for informed medical advice
from your own physician.