I think about this when I see the patient that is my most typical new patient referral. A man who has undergone a radical prostatectomy or surgery, and now has a rising PSA.
My approach to discussing prognosis with these patients is to talk about the ‘rule of thirds’ which is to say that roughly one third of those who experience a serologic relapse may not need any further anti-cancer treatment, one third will require treatment to prevent metastases, but may live to or past the average life expectancy for their age but not die of the disease. The final one third of those who ‘relapse’ will face a life threatening form of prostate cancer. I don’t have data on whether it actually breaks out cleanly as one third-one third-one third, but you get the idea.
How do we know which is which? Because PSA is the only thing we have to measure at these points, it makes sense that we would focus on the dynamics of this biomarker in order to make disease assessments. The most useful biomarker post-relapse is not the level of the PSA itself but rather the rate of rise of PSA, quantified as the PSA doubling time. It is expressed in months.
I wrote a paper several years ago on how we should utilize these data in the clinic and now, about 10 years later, my approach is only a little bit different.
Here’s how I think through it:
1. PSA doubling time < 3 months. This is the worst prognosis and, more importantly, the data show that patients with a PSADT < 3 months are more likely to die of prostate cancer cause than any other cause. It is, to borrow the word from Anthony D’Amico’s papers of a decade ago, a “surrogate” for prostate cancer death[1, 2]. Thus, in my opinion, these patients need treatment and, most likely will benefit from early treatment.
2. PSA Doubling time >12 months. These are patients with a relatively good prognosis and were the comparator group in the D’Amico studies analyses. Most of them are not going to die of prostate cancer. Thus, maybe these are the ones who can be observed. I’ve followed many of these patients over the years, and have followed them for years. I won’t say that they don’t develop metastases, because I have seen it – in the nodes and even in the lung, but prostate cancer death in these patients is rare.
And everybody in the middle is in the middle. I think of it as a continuous variable. The shorter the PSADT, the worse the prognosis, but 12 months is a pretty convenient cutoff above which to say, this may not need be their biggest worry.
But, with the fact that there is some tumor in there established we are left asking the same question: where’s the cancer? I get this question a lot. Heck, I ask this question a lot.
I routinely go through a few explanations: Its in liquid form, its in the bone marrow, its maybe in the prostate itself in a post radiation patient, or its hidden in the lymph nodes. Or, its in some combination of these anatomic areas.
Recently, at the annual Prostate Cancer Foundation Retreat in Washington D.C., Ken Pienta, highly regarded as one of the deeper thinkers on the topic of tumor micro-environment and metastases generally, reviewed the data and his proposed theories on this process. It is hard to observe a cell leaving the primary and traveling to a metastatic site, but we do know that the circulation will frequently harbor ample supplies of circulating tumor cells (CTC’s) as well as cells that have homed in on the bone marrow, and are assessable through a bone marrow biopsy or aspirate – these are called disseminated tumor cells, or DTC’s.
In the setting of localized prostate cancer, when supposedly the primary tumor is ‘shedding’ the metastatic seedlings, DTC’s are not that common – only single digits worth of them have been counted when exploring about 100 bone marrows. Yet, in the studies that have been done many of these patients had low PSA values so there may not be a dominant clone in any kind of a high quantity that is being measured.
Pienta’s group  has published some nice work and a conceptual framework for all of this (Molecular Cancer Research 2017). In this very compelling review paper, they borrow a concept from environmental ecology known as the Optimal Foraging Strategy.
As the name implies, the cancers that are able to spread and be deadly are those that can forage most effectively. This concept, however, rests on several key observations. The first is that cancer cells can move within the primary tumor. The second is that the fittest members of a foraging population will adapt to changes in environmental resources like oxygen and glucose, reduced environmental threats of predation from T cells and macrophages and will have optimized movement ability.
De Groot and colleagues state that while most cancer cells are stationary foragers - they stay in place and wait for the resources to come to them, a select few are mobile foragers. Both foraging strategies exist within the primary tumor , those that obtain mobility are more adapted. The very rare cells that are capable of movement, which may be those capable of epithelial–mesenchymal transition (EMT) may be those that take on mobile forager status. Cells will migrate to areas of richer oxygen concentrations within a hypoxic primary tumor. They label the exit routes of tumor cells dispersal corridors, which we know to be blood vessels, lymph channels and nerves. Each has their own advantage and disadvantage. Blood vessels offer speed and are rich in oxygen and glucose, but is a high volume corridor – leaving the cancer cell like a tiny fish in a big ocean, often with threats such as the shear stress of heart valves and vessel walls. A lymph node is a high predation environment due to the high concentration of T cells, but offers flow and the ability of tumor cells to become concentrated, a nerve is a low flow environment but requires movement potential and thus a cell dispersing through a nerve channel is likely to be a highly adapted, ‘tough’ tumor cell that is capable of movement.
They did some remarkable, if theoretical, calculations. If, at any given time, a patient harbors 1 CTC per 1 ml of blood, then there are only 5000 CTCs in the entire blood pool (estimating 5L of blood). Assuming a CTC circulates 1 time then 'lands' somewhere or dies, then 2.6x109 CTCs are shed from a primary tumor in a year based on a cardiac output of 5L and 60 beats per minute. Then they considered the fate of all those shed cells, and observed that observational data has shown that typically 1 disseminated tumor cell may be identified in 1mL of bone marrow. Given a 1.75L marrow volume, there may only be about 1750 DTC present outside of the prostate. Calculating that the overall "efficiency" of CTC to produce 1 bone marrow DTC in 10 years is 1 DTC to every 15x106 total CTC.
Finally, they calculated the efficiency of the CTC to produce a clinical metastasis. If, of 1750 DTC in the total marrow compartment, only 0.01% is 'reawakened' to become clinically active they calculate that the overall efficiency of CTCs to produce 1 clinical bone met is 1 metastasis per every 1.44x109 total CTC. That’s one metastasis for every 1.44 billion shed cancer cells. And most cells aren’t shed.
So maybe that’s where it is. Rare cells that enter the circulation hide out for a while in the marrow and MAY OR MAY NOT become active and those are the ones that form the metastatic tumors that kill men with this disease.
What does it mean in the clinic? To me it points to where other therapeutic data such as Latitude, Stampede and especially the neoadjuvant studies that have been done do: early systemic therapy - typically in the form of effective hormonal approaches, and maybe even chemotherapy, may be the key to a) turning off the ‘fountain’ of cancer cells coming out of the primary tumor – 1.6 billion per year and b) making it hard for them to find succor in a testosterone and oxygen rich environment.
We are awaiting the results of CALGB 90203 a study of approximately 1000 patients with high risk localized disease who were randomized to receive ADT with or without docetaxel prior to undergoing a radical prostatectomy. Radiation plus chemotherapy studies look promising but need more maturation. If positive, we’ll likely be crediting the chemotherapy not only with treating the metastases, but also for impeding the egress of the cancer out of the primary by impairing the microtubules.
Written by: Charles Ryan, MD
1. D'Amico AV, Cote K, Loffredo M et al. Determinants of prostate cancer-specific survival after radiation therapy for patients with clinically localized prostate cancer. J Clin Oncol 2002; 20: 4567-4573.
2. D'Amico AV, Moul J, Carroll PR et al. Prostate specific antigen doubling time as a surrogate end point for prostate cancer specific mortality following radical prostatectomy or radiation therapy. J Urol 2004; 172: S42-46; discussion S46-47.
3. de Groot AE, Roy S, Brown JS et al. Revisiting Seed and Soil: Examining the Primary Tumor and Cancer Cell Foraging in Metastasis. Mol Cancer Res 2017; 15: 361-370.
Long-term follow-up of a phase II trial of chemotherapy plus hormone therapy for biochemical relapse after definitive local therapy for prostate cancer - Abstract
Watch: Impacts of Latitude and Stampede: Urologist and Urologic Oncologist's Perspective