What would you do -- if anything?

Best left to TA and others more knowledgeable. But looks like a trend to me. Albeit a slow one

The definition of a reoccurrance is nadir +2, so a psa of 3.96 would be officially a reoccurrance.
The rate of rise at a consistent rate over 15 months is not a bounce and is acting as a growing tumor would act, so this alone is disconcerting.
I'de be looking for an oncologist specializing in PC to guide you.

JT

Mary, I agree with the others that this does not look like a bounce. I agree also that you should seek Dr. Kumrava's input as to what he thinks is going on.

For a list of medical oncologists who specialize in prostate cancer, please see that appendix of Dr. Scholz's book, "Invasion of the Prostate Snatchers".

Good luck.

Jim

Hi Mary,

Dr King knows best (always my motto). He has seen a lot of this sort of thing. I persist in the belief that John has a stubborn case of prostatitis, and I'm sure that the catheters didn't help. The next investigative step would be a targeted biopsy, but that might only serve to aggravate the prostatitis. I'd also remind you that a biochemical failure is just an arbitrary cutoff that says that there may be something worth investigating, and it is very different from a clinical failure. I don't know if a DRE would tell you anything - the prostatitis sometimes makes the prostate feel inflamed to the touch, but often it doesn't - especially when PSA is this low.

You may have heard that Dr. Kamrava has moved to Cedars-Sinai and Dr. Demanes has come out of retirement to fill in at UCLA. In fact, I just met with him and a patient yesterday.

This would be so much easier if John's prostate would just cooperate and his PSA would just follow a smooth downward curve. Frustrating, I know.

Glad you contacted Dr. King. He is so pleasant and really does a nice job in responding to emails: quickly and on point!

Mel

I have no idea why you want to see an oncologist. Do you think an oncologist will know anything about PSA after radiation?

John can certainly try a round of antibiotics. Did he ever take them for prostatitis before? If so, you may want to switch to a different one. Usually, they don't like to give antibiotics for this for 2 reasons: (1) it usually has no effect on prostatitis yet creates an expectation it will, and (2) it creates more resistance in the future. On the upside, if it does happen to drop his PSA, it will help relieve your anxiety. But if it doesn't, will it make you more anxious?

If you want more stats from medical journals, they sometimes publish some of them in the appendix. You're right that there are conventions they use that one gets used to after seeing a lot of them on a subject. You don't often see regression coefficients. More commonly, they show the OR, and whether it is statistically significant. That captures the relative effect on a normalized basis. There's a lot of "behind the scenes" corrections made for confounding variables. Sometimes they tell you which statistical package they used and which techniques in that package was used in their calculations.

It's one of the benefits of surgery - if PSA remains detectable, you know immediately that something's amiss. Radiation requires patience.

I have a folder on the topic I label "bad research." I'll email its contents to you. Peer review certainly helps overcome some of the flaws, but I'm surprised that peer review sometimes is lackadaisical. I think all journals should reveal the identities of peer reviewers.

We make a distinction between clinical significance and statistical significance. An OR of 0.99 may be statistically significant on a large sample size, but not meaningful enough to alter treatments. This came up in a study I was just writing up. To the authors' credit, they reported both the minimal clinically-detectable differences (which in this case was half a std deviation) and the significantly detected differences (which was at least twice the minimum). When I look at p-values, it's a minimum qualification. But you have to look at both the univariate and multivariate statistics because variables are seldom independent. But those stats are just the tip of the iceberg. Levels of evidence are important in medicine, as are the qualitative grades assigned (by "experts") to the evidence. So much of the evidence we have is low level and poor quality - yet what are we to do if that's the best we've got? That's where the patient/clinician's judgment comes in.